Open source
Open source
Open source
If the essays speak the judgment aloud, the repositories are where that judgment is handed over to the world in workable form. Each project page also adds a site-native reading of the README, so the repos can be read as products, workflows, and public proof.
Repository quick routes
Scan a one-screen row index before opening deeper dossier cards and proof atlas entries.
Repository activity lanes
Audit repository cadence by recent push date, then route each row back into topic lanes, proof paths, forum lanes, and collaboration.
12 activity rows
Proof continuity matrix
Audit how each repository connects into topic lanes, cluster proof paths, forum lanes, and collaboration routes.
10 continuity rows
Topic proof lanes
Group repository proof by topic lanes, then route each lane into longform, forum, and collaboration paths.
6 topic lanes
01 / Technical Distribution
9 repos / 8.4K stars
How open-source proof, X field notes, case studies, SEO, and founder outreach become one trust system.
02 / Harness Engineering
5 repos / 7.9K stars
The operating layer around models: context policy, constraints, evals, memory, approval boundaries, and cleanup loops.
03 / Tooling
9 repos / 755 stars
Software that makes technical thought less fragile and public work easier to sustain.
04 / AI-Native Workbenches
11 repos / 415 stars
Composable skills, local workflows, agent utilities, and the product surfaces that make AI work repeatable.
05 / Memory
5 repos / 96 stars
Persistence, compression, retrieval, and the editorial judgment behind what a system should keep.
06 / Agents
3 repos / 8 stars
Delegation, orchestration, recovery, and what breaks when multi-step intelligence is treated as magic.
Proof atlas index
Scan the public proof system by repository, topic, continuation route, and collaboration edge.
10 indexed repos
01 / Python / 7.9k stars
Repository / yizhiyanhua-ai
fireworks-tech-graph
Hand-made one-off diagrams that collapse after a single meeting or slide deck.
Dossier page / fireworks-tech-graph
Dossier page
Harness Engineering
Continue into the topic archive that frames this repository.
Topic archive / Harness Engineering
Topic route
Harness engineering for agentic AI products
Place this repository back into the founder question around Harness engineering for agentic AI products.
Cluster route / AI Thesis
Cluster route
Founder collaboration edge
Continue repository proof into founding engineer, design partner, and contract build paths.
Fit route / founder collaboration
Fit route
02 / Python / 94 stars
Repository / yizhiyanhua-ai
fireworks-skill-memory
One giant memory dump that cannot distinguish a reusable pattern from random residue.
Dossier page / fireworks-skill-memory
Dossier page
Memory
Continue into the topic archive that frames this repository.
Topic archive / Memory
Topic route
Wider proof continuation
Return to the thesis proof graph when no cluster is strongly matched.
Cluster route / thesis map
Cluster route
Founder collaboration edge
Continue repository proof into founding engineer, design partner, and contract build paths.
Fit route / founder collaboration
Fit route
03 / Python / 8 stars
Repository / yizhiyanhua-ai
fireworks-sessions-saver
Fragile long-running sessions where useful context disappears after pauses, crashes, or handoffs.
Dossier page / fireworks-sessions-saver
Dossier page
Agents
Continue into the topic archive that frames this repository.
Topic archive / Agents
Topic route
Harness engineering for agentic AI products
Place this repository back into the founder question around Harness engineering for agentic AI products.
Cluster route / AI Thesis
Cluster route
Founder collaboration edge
Continue repository proof into founding engineer, design partner, and contract build paths.
Fit route / founder collaboration
Fit route
04 / TypeScript / 2 stars
Repository / yizhiyanhua-ai
paper-pal
PDF hoarding and one-pass reading that never turns into reusable understanding.
Dossier page / paper-pal
Dossier page
Tooling
Continue into the topic archive that frames this repository.
Topic archive / Tooling
Topic route
Wider proof continuation
Return to the thesis proof graph when no cluster is strongly matched.
Cluster route / thesis map
Cluster route
Founder collaboration edge
Continue repository proof into founding engineer, design partner, and contract build paths.
Fit route / founder collaboration
Fit route
05 / Python / 390 stars
Repository / yizhiyanhua-ai
media-downloader
Manual downloading, clipping, and asset chasing across websites and apps.
Dossier page / media-downloader
Dossier page
Tooling
Continue into the topic archive that frames this repository.
Topic archive / Tooling
Topic route
Technical distribution for AI builders
Place this repository back into the founder question around Technical distribution for AI builders.
Cluster route / AI Thesis
Cluster route
Founder collaboration edge
Continue repository proof into founding engineer, design partner, and contract build paths.
Fit route / founder collaboration
Fit route
06 / Python / 165 stars
Repository / yizhiyanhua-ai
skills-updater
Stale skill libraries that quietly drift out of date after initial excitement.
Dossier page / skills-updater
Dossier page
Tooling
Continue into the topic archive that frames this repository.
Topic archive / Tooling
Topic route
Wider proof continuation
Return to the thesis proof graph when no cluster is strongly matched.
Cluster route / thesis map
Cluster route
Founder collaboration edge
Continue repository proof into founding engineer, design partner, and contract build paths.
Fit route / founder collaboration
Fit route
07 / Shell / 91 stars
Repository / yizhiyanhua-ai
chuinb-skill
Unstructured assistance that never becomes a crisp reusable behavior.
Dossier page / chuinb-skill
Dossier page
Tooling
Continue into the topic archive that frames this repository.
Topic archive / Tooling
Topic route
Wider proof continuation
Return to the thesis proof graph when no cluster is strongly matched.
Cluster route / thesis map
Cluster route
Founder collaboration edge
Continue repository proof into founding engineer, design partner, and contract build paths.
Fit route / founder collaboration
Fit route
08 / Python / 51 stars
Repository / yizhiyanhua-ai
youtube-ai-digest
Manual browsing across YouTube channels with weak filtering and no synthesis.
Dossier page / youtube-ai-digest
Dossier page
Tooling
Continue into the topic archive that frames this repository.
Topic archive / Tooling
Topic route
Technical distribution for AI builders
Place this repository back into the founder question around Technical distribution for AI builders.
Cluster route / AI Thesis
Cluster route
Founder collaboration edge
Continue repository proof into founding engineer, design partner, and contract build paths.
Fit route / founder collaboration
Fit route
09 / Python / 51 stars
Repository / yizhiyanhua-ai
zimage-skill
Manual image prompting with no durable interface back into broader systems.
Dossier page / zimage-skill
Dossier page
Tooling
Continue into the topic archive that frames this repository.
Topic archive / Tooling
Topic route
Wider proof continuation
Return to the thesis proof graph when no cluster is strongly matched.
Cluster route / thesis map
Cluster route
Founder collaboration edge
Continue repository proof into founding engineer, design partner, and contract build paths.
Fit route / founder collaboration
Fit route
10 / HTML / 39 stars
Repository / yizhiyanhua-ai
fireworks-design
🎆 A Claude Code workflow that replicates ClaudeDesign — parallel design exploration, panel judging, synthesis, and adversarial refinement to a single world-class frontend page. 双语 README。
Dossier page / fireworks-design
Dossier page
Thesis map
Route back to the thesis map when no explicit topic is attached.
Thesis route / proof graph
Thesis route
Wider proof continuation
Return to the thesis proof graph when no cluster is strongly matched.
Cluster route / thesis map
Cluster route
Founder collaboration edge
Continue repository proof into founding engineer, design partner, and contract build paths.
Fit route / founder collaboration
Fit route
Featured repositories
Begin with the works carrying the most weight.
Begin with the works carrying the most weight.
fireworks-tech-graph
Technical diagrams as durable product and distribution infrastructure.
This repo is market proof that the builder can turn vague technical structure into public artifacts developers want to share.
Explain AI systems faster to users, teammates, investors, and customers.
Build a repeatable visual grammar around complex workflows and infrastructure.
Topic continuation
Technical Distribution
Continue into the topic archive behind this project.
Topic archive / Technical Distribution
Topic route
Topic continuation
Harness Engineering
Continue into the topic archive behind this project.
Topic archive / Harness Engineering
Topic route
Cluster continuation
Cluster 01
What has to surround the model before an AI workflow becomes a product?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Cluster continuation
Cluster 06
How does technical work compound into trust, search traffic, and serious conversations?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Dossier surface
Site-native README reading and proof ledger
Start with the site-native dossier before deciding whether to continue into GitHub or the case study.
Dossier page / fireworks-tech-graph
Dossier page
Repository source
Original GitHub repository
Return to the raw source, issues, commits, and the external public context.
Repository / yizhiyanhua-ai/fireworks-tech-graph
Repository route
Live surface
Public site or runnable surface
If the project has a public surface, enter its live presentation directly from here.
Live route / public project surface
Live route
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / fireworks-tech-graph
Dossier page
Case
Founder-readable flagship case study
This project also has a founder-readable case study to continue into.
Case page / fireworks-tech-graph
Case route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
fireworks-skill-memory
Per-skill memory that turns past work into reusable operating context.
This is one of the clearest proofs on the site that AI memory should be structured, selective, and recoverable rather than merely larger.
Create lasting memory for repeated AI skills without overloading every session.
Turn operator learning into a real asset the system can re-enter later.
Topic continuation
Memory
Continue into the topic archive behind this project.
Topic archive / Memory
Topic route
Topic continuation
AI-Native Workbenches
Continue into the topic archive behind this project.
Topic archive / AI-Native Workbenches
Topic route
Dossier surface
Site-native README reading and proof ledger
Start with the site-native dossier before deciding whether to continue into GitHub or the case study.
Dossier page / fireworks-skill-memory
Dossier page
Repository source
Original GitHub repository
Return to the raw source, issues, commits, and the external public context.
Repository / yizhiyanhua-ai/fireworks-skill-memory
Repository route
Live surface
No separate public surface yet
For now the strongest public surface is still the dossier, README, and proof continuation chain.
Use dossier / repository / proof routes
Library surface
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / fireworks-skill-memory
Dossier page
Thesis
Wider thesis and proof graph
Return to the wider proof map.
Thesis route / proof graph
Thesis route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
fireworks-sessions-saver
Checkpointing for coding sessions that will inevitably be interrupted.
Interruption is not an edge case in real AI work. A team that can resume cleanly has a direct execution advantage.
Design recovery paths for long-running agentic workflows.
Reduce the hidden cost of context loss across team handoff.
Topic continuation
Agents
Continue into the topic archive behind this project.
Topic archive / Agents
Topic route
Topic continuation
Harness Engineering
Continue into the topic archive behind this project.
Topic archive / Harness Engineering
Topic route
Cluster continuation
Cluster 01
What has to surround the model before an AI workflow becomes a product?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Dossier surface
Site-native README reading and proof ledger
Start with the site-native dossier before deciding whether to continue into GitHub or the case study.
Dossier page / fireworks-sessions-saver
Dossier page
Repository source
Original GitHub repository
Return to the raw source, issues, commits, and the external public context.
Repository / yizhiyanhua-ai/fireworks-sessions-saver
Repository route
Live surface
No separate public surface yet
For now the strongest public surface is still the dossier, README, and proof continuation chain.
Use dossier / repository / proof routes
Library surface
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / fireworks-sessions-saver
Dossier page
Cluster
Harness engineering for agentic AI products thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Repository shelf
The remaining repositories continue on the side shelf.
The remaining repositories continue on the side shelf.
TypeScript / 2 stars / yizhiyanhua-ai
paper-pal
Research reading as an active workflow of return, notes, and synthesis.
PDF hoarding and one-pass reading that never turns into reusable understanding.
Serious AI work depends on reading systems, not just browsing papers. This project treats comprehension as production work.
Topic continuation
Tooling
Continue into the topic archive tied to this project.
Topic archive / Tooling
Topic route
Topic continuation
AI-Native Workbenches
Continue into the topic archive tied to this project.
Topic archive / AI-Native Workbenches
Topic route
Build better research loops inside AI teams.
Turn technical reading into a durable internal knowledge asset.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / paper-pal
Dossier page
Thesis
Wider thesis and proof graph
Return to the wider proof map.
Thesis route / proof graph
Thesis route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 390 stars / yizhiyanhua-ai
media-downloader
A utility layer for turning media gathering into a repeatable workflow.
Manual downloading, clipping, and asset chasing across websites and apps.
Creative and technical distribution systems both break when asset collection stays manual and scattered.
Topic continuation
Technical Distribution
Continue into the topic archive tied to this project.
Topic archive / Technical Distribution
Topic route
Topic continuation
Tooling
Continue into the topic archive tied to this project.
Topic archive / Tooling
Topic route
Cluster continuation
Cluster 06
How does technical work compound into trust, search traffic, and serious conversations?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Speed up content operations around AI products and public demos.
Reduce repetitive asset preparation work for small teams.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / media-downloader
Dossier page
Cluster
Technical distribution for AI builders thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 165 stars / yizhiyanhua-ai
skills-updater
Maintenance infrastructure for living skill ecosystems.
Stale skill libraries that quietly drift out of date after initial excitement.
Reusable AI workflows decay unless updates and discovery stay easy. Maintenance tooling is what keeps reuse real.
Topic continuation
AI-Native Workbenches
Continue into the topic archive tied to this project.
Topic archive / AI-Native Workbenches
Topic route
Topic continuation
Tooling
Continue into the topic archive tied to this project.
Topic archive / Tooling
Topic route
Keep team AI workflows fresh without rebuilding them from scratch.
Design update loops and discovery mechanics for internal AI tooling.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / skills-updater
Dossier page
Thesis
Wider thesis and proof graph
Return to the wider proof map.
Thesis route / proof graph
Thesis route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Shell / 91 stars / yizhiyanhua-ai
chuinb-skill
A narrow skill as evidence of taste in workflow packaging.
Unstructured assistance that never becomes a crisp reusable behavior.
Small skills show whether the builder can choose a boundary well. That judgment matters more than raw feature count.
Topic continuation
AI-Native Workbenches
Continue into the topic archive tied to this project.
Topic archive / AI-Native Workbenches
Topic route
Topic continuation
Tooling
Continue into the topic archive tied to this project.
Topic archive / Tooling
Topic route
Test how small AI behaviors can become product surfaces.
Evaluate packaging discipline on a small, inspectable scope.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / chuinb-skill
Dossier page
Thesis
Wider thesis and proof graph
Return to the wider proof map.
Thesis route / proof graph
Thesis route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 51 stars / yizhiyanhua-ai
youtube-ai-digest
A workflow for turning noisy AI video streams into digestible signals.
Manual browsing across YouTube channels with weak filtering and no synthesis.
A lot of market and technical intelligence now appears in video first. Digest tooling helps founders keep up without losing whole days.
Topic continuation
Technical Distribution
Continue into the topic archive tied to this project.
Topic archive / Technical Distribution
Topic route
Topic continuation
Tooling
Continue into the topic archive tied to this project.
Topic archive / Tooling
Topic route
Cluster continuation
Cluster 06
How does technical work compound into trust, search traffic, and serious conversations?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Build faster market-intelligence loops around AI trends and launches.
Convert long-form video information into operator-ready notes.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / youtube-ai-digest
Dossier page
Cluster
Technical distribution for AI builders thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 51 stars / yizhiyanhua-ai
zimage-skill
Image generation as a reusable workflow component, not a one-off trick.
Manual image prompting with no durable interface back into broader systems.
Multimodal products need image operations that are structured, callable, and easy to embed in wider workflows.
Topic continuation
AI-Native Workbenches
Continue into the topic archive tied to this project.
Topic archive / AI-Native Workbenches
Topic route
Topic continuation
Tooling
Continue into the topic archive tied to this project.
Topic archive / Tooling
Topic route
Add image generation to AI workflows without reinventing the whole surface each time.
Treat multimodal operations as product building blocks.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / zimage-skill
Dossier page
Thesis
Wider thesis and proof graph
Return to the wider proof map.
Thesis route / proof graph
Thesis route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
HTML / 39 stars / yizhiyanhua-ai
fireworks-design
🎆 A Claude Code workflow that replicates ClaudeDesign — parallel design exploration, panel judging, synthesis, and adversarial refinement to a single world-class frontend page. 双语 README。
🎆 A Claude Code workflow that replicates ClaudeDesign — parallel design exploration, panel judging, synthesis, and adversarial refinement to a single world-class frontend page. 双语 README。
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / fireworks-design
Dossier page
Thesis
Wider thesis and proof graph
Return to the wider proof map.
Thesis route / proof graph
Thesis route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 5 stars / yizhiyanhua-ai
agent-skills
A skill library for turning repeated prompts into reusable operating units.
Prompt copy-paste culture with weak reuse and no durable operating pattern.
AI teams win when repeated work becomes packaged behavior. Skills make that packaging explicit.
Topic continuation
AI-Native Workbenches
Continue into the topic archive tied to this project.
Topic archive / AI-Native Workbenches
Topic route
Topic continuation
Tooling
Continue into the topic archive tied to this project.
Topic archive / Tooling
Topic route
Create repeatable internal workflows around common AI tasks.
Package operator know-how into team-ready units.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / agent-skills
Dossier page
Thesis
Wider thesis and proof graph
Return to the wider proof map.
Thesis route / proof graph
Thesis route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 3 stars / yizhiyanhua-ai
fireworks-macapp-creator
A bridge from agent workflows into packaged desktop software.
CLI-only experiments that never cross into installable user-facing products.
A lot of AI utility dies as scripts. Product value appears when those utilities become distributable surfaces people can actually run.
Topic continuation
AI-Native Workbenches
Continue into the topic archive tied to this project.
Topic archive / AI-Native Workbenches
Topic route
Topic continuation
Technical Distribution
Continue into the topic archive tied to this project.
Topic archive / Technical Distribution
Topic route
Cluster continuation
Cluster 06
How does technical work compound into trust, search traffic, and serious conversations?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Turn internal agent workflows into productized desktop experiences.
Explore packaging and distribution options beyond scripts.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / fireworks-macapp-creator
Dossier page
Cluster
Technical distribution for AI builders thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 2 stars / yizhiyanhua-ai
skill-knowledge-extractor
A tool for turning tacit workflow know-how into extractable assets.
Knowledge staying trapped inside long chats, docs, and operator memory.
The valuable part of AI work is often implicit. Extraction tools make that hidden layer easier to package and reuse.
Topic continuation
AI-Native Workbenches
Continue into the topic archive tied to this project.
Topic archive / AI-Native Workbenches
Topic route
Topic continuation
Memory
Continue into the topic archive tied to this project.
Topic archive / Memory
Topic route
Capture workflow knowledge before it disappears into chat history.
Create cleaner inputs for building skills, docs, and internal playbooks.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / skill-knowledge-extractor
Dossier page
Thesis
Wider thesis and proof graph
Return to the wider proof map.
Thesis route / proof graph
Thesis route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Shell / 1 stars / yizhiyanhua-ai
fireworks-radio
Ambient audio as an AI-native working surface.
Manual context switching between coding and separate media tools for focus management.
AI products do not all need to look like dashboards. Some create atmosphere, pacing, and sustained attention around technical work.
Topic continuation
AI-Native Workbenches
Continue into the topic archive tied to this project.
Topic archive / AI-Native Workbenches
Topic route
Topic continuation
Technical Distribution
Continue into the topic archive tied to this project.
Topic archive / Technical Distribution
Topic route
Cluster continuation
Cluster 06
How does technical work compound into trust, search traffic, and serious conversations?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Experiment with non-text product surfaces around AI work.
Think about developer experience beyond prompts and forms.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / fireworks-radio
Dossier page
Cluster
Technical distribution for AI builders thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 1 stars / yizhiyanhua-ai
skill-adoption-planner
A planning layer for whether a skill should actually be adopted.
Collecting skills without evaluating fit, payoff, or operational friction.
AI workflow libraries fail when everything gets added but nothing gets adopted. Adoption discipline is part of product quality.
Topic continuation
AI-Native Workbenches
Continue into the topic archive tied to this project.
Topic archive / AI-Native Workbenches
Topic route
Topic continuation
Technical Distribution
Continue into the topic archive tied to this project.
Topic archive / Technical Distribution
Topic route
Cluster continuation
Cluster 06
How does technical work compound into trust, search traffic, and serious conversations?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Prioritize which AI workflows deserve rollout in a real team.
Bring product and operations thinking into skill adoption.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / skill-adoption-planner
Dossier page
Cluster
Technical distribution for AI builders thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 0 stars / yizhiyanhua-ai
fireworks-ai-tools-memory
Memory for tool-using AI systems that need continuity across work.
Fresh-start tool usage where the system forgets prior patterns and operator preferences each session.
Tool memory is where AI products start feeling less stateless and more like real operating systems.
Topic continuation
Memory
Continue into the topic archive tied to this project.
Topic archive / Memory
Topic route
Topic continuation
Harness Engineering
Continue into the topic archive tied to this project.
Topic archive / Harness Engineering
Topic route
Cluster continuation
Cluster 01
What has to surround the model before an AI workflow becomes a product?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Preserve tool history, preferences, and state around repeated AI workflows.
Push agent products toward continuity without pretending context never breaks.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / fireworks-ai-tools-memory
Dossier page
Cluster
Harness engineering for agentic AI products thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 0 stars / ccc7574
agent-architect-lab
A public lab for agent-system structure, not just agent demos.
One-shot prototype building without a reusable architecture vocabulary.
The market has many agent surfaces but fewer builders who can reason about boundaries, orchestration, and failure modes before shipping.
Topic continuation
Agents
Continue into the topic archive tied to this project.
Topic archive / Agents
Topic route
Topic continuation
Harness Engineering
Continue into the topic archive tied to this project.
Topic archive / Harness Engineering
Topic route
Cluster continuation
Cluster 01
What has to surround the model before an AI workflow becomes a product?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Explore architecture options before a team hardens the wrong workflow.
Frame agent products as systems with ownership and recovery paths.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / agent-architect-lab
Dossier page
Cluster
Harness engineering for agentic AI products thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 0 stars / ccc7574
kv-agent-lab
A lab for key-value state and durable memory patterns in agent systems.
Overcomplicated memory abstractions before basic state primitives are understood.
Durability often depends on state design more than model choice. This project keeps attention on that lower layer.
Topic continuation
Memory
Continue into the topic archive tied to this project.
Topic archive / Memory
Topic route
Topic continuation
Harness Engineering
Continue into the topic archive tied to this project.
Topic archive / Harness Engineering
Topic route
Cluster continuation
Cluster 01
What has to surround the model before an AI workflow becomes a product?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Explore durable state patterns for agents before betting on a heavier architecture.
Clarify where persistence should live in an AI workflow.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / kv-agent-lab
Dossier page
Cluster
Harness engineering for agentic AI products thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 0 stars / yizhiyanhua-ai
agent-shared-memory-demo
A concrete window into shared memory for multi-agent work.
Abstract discussion about collaboration without inspectable shared state.
Shared memory is where collaboration either becomes leverage or confusion. Even a demo is valuable when it clarifies coordination shape.
Topic continuation
Memory
Continue into the topic archive tied to this project.
Topic archive / Memory
Topic route
Topic continuation
Agents
Continue into the topic archive tied to this project.
Topic archive / Agents
Topic route
Prototype how multiple agents should share evidence and task memory.
Make memory policy visible before deeper productization.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / agent-shared-memory-demo
Dossier page
Thesis
Wider thesis and proof graph
Return to the wider proof map.
Thesis route / proof graph
Thesis route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 0 stars / yizhiyanhua-ai
skill-roi-calculator
An ROI lens for deciding whether an AI skill is worth operationalizing.
Shipping AI utilities with no model for time saved, effort required, or adoption payoff.
The next wave of AI work needs economic judgment, not just enthusiasm. ROI tooling makes that judgment explicit.
Topic continuation
Technical Distribution
Continue into the topic archive tied to this project.
Topic archive / Technical Distribution
Topic route
Topic continuation
AI-Native Workbenches
Continue into the topic archive tied to this project.
Topic archive / AI-Native Workbenches
Topic route
Cluster continuation
Cluster 06
How does technical work compound into trust, search traffic, and serious conversations?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Evaluate whether a workflow deserves to become a product or internal standard.
Add commercial discipline to AI workflow design.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / skill-roi-calculator
Dossier page
Cluster
Technical distribution for AI builders thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Python / 0 stars / yizhiyanhua-ai
rhythm-master
A playful interface that still proves interaction design range.
Purely utilitarian engineering output with no evidence of interaction feel.
Not every proof asset has to be infra-heavy. Sometimes a small interaction system shows timing, feedback, and product sensitivity.
Topic continuation
Tooling
Continue into the topic archive tied to this project.
Topic archive / Tooling
Topic route
Topic continuation
Technical Distribution
Continue into the topic archive tied to this project.
Topic archive / Technical Distribution
Topic route
Cluster continuation
Cluster 06
How does technical work compound into trust, search traffic, and serious conversations?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Show that the builder can shape interface feedback, not just backend logic.
Add breadth to the public proof system around product feel.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / rhythm-master
Dossier page
Cluster
Technical distribution for AI builders thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Markdown / 0 stars / yizhiyanhua-ai
.github
The operating surface for the whole organization.
Ad hoc org management spread across memory, chat, and one-off repository setup.
Serious open-source work compounds through defaults: templates, issue flows, contribution rules, and public coordination mechanics.
Topic continuation
Technical Distribution
Continue into the topic archive tied to this project.
Topic archive / Technical Distribution
Topic route
Topic continuation
Tooling
Continue into the topic archive tied to this project.
Topic archive / Tooling
Topic route
Cluster continuation
Cluster 06
How does technical work compound into trust, search traffic, and serious conversations?
Start at AI Thesis, then continue through the thesis-side route.
Lead proof / AI Thesis
Cluster route
Design cleaner public operating systems around multiple repos.
Improve contributor trust and onboarding quality.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / .github
Dossier page
Cluster
Technical distribution for AI builders thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Shelf routes
Open thesisFlagship proof
Calibrate the proof bar through the flagship case
Use the strongest public case to calibrate the quality bar for the whole repository hall.
Case page / fireworks-tech-graph
Case route
Commercial route
Turn repository proof into a working conversation
Let repository signal flow naturally into founding engineer, design partner, and contract build paths.
Fit route / founder collaboration
Fit route
Repository routes
Repositories are not endpoints. They are proof nodes with onward routes.
Repositories are not endpoints. They are proof nodes with onward routes.
Proof clusters
The project library and thesis clusters sit on the same proof map.
The project library and thesis clusters sit on the same proof map.
01 / Cluster 01
3 routesHarness engineering for agentic AI products
What has to surround the model before an AI workflow becomes a product?
Repository shelf
fireworks-tech-graph leads this cluster
2 repository nodes support this cluster.
Proof library / repository shelf
Lead proof
AI Thesis
Enter this founder question through AI Thesis first.
Cluster route / AI Thesis
Commercial edge
Founder collaboration routes
Continue this cluster's public proof into founder collaboration.
Fit route / founder collaboration
Lead repository
fireworks-tech-graph
Hand-made one-off diagrams that collapse after a single meeting or slide deck.
Dossier page / fireworks-tech-graph
Dossier page
Lead dossier
Start with the strongest repository dossier inside this cluster.
Dossier page / fireworks-tech-graph
Dossier page
Cluster route
Route the repository proof back into the thesis and founder question.
Cluster route / AI Thesis
Cluster route
Fit route
Turn cluster-level proof into a concrete working conversation.
Fit route / founder collaboration
Fit route
02 / Cluster 02
3 routesLong-running agents and resumable work
Where does the user lose trust when an agent runs for hours?
Repository shelf
Enter this cluster through the repository library
0 repository nodes support this cluster.
Proof library / repository shelf
Lead proof
AI Thesis
Enter this founder question through AI Thesis first.
Cluster route / AI Thesis
Commercial edge
Founder collaboration routes
Continue this cluster's public proof into founder collaboration.
Fit route / founder collaboration
Lead dossier
Enter the cluster through the broader proof library.
Dossier page / repository library
Dossier page
Cluster route
Route the repository proof back into the thesis and founder question.
Cluster route / AI Thesis
Cluster route
Fit route
Turn cluster-level proof into a concrete working conversation.
Fit route / founder collaboration
Fit route
03 / Cluster 03
3 routesDeveloper communication for AI infrastructure
Can someone understand the product before they understand the whole codebase?
Repository shelf
Enter this cluster through the repository library
0 repository nodes support this cluster.
Proof library / repository shelf
Lead proof
fireworks-tech-graph case study
Enter this founder question through fireworks-tech-graph case study first.
Cluster route / fireworks-tech-graph case study
Commercial edge
Founder collaboration routes
Continue this cluster's public proof into founder collaboration.
Fit route / founder collaboration
Lead dossier
Enter the cluster through the broader proof library.
Dossier page / repository library
Dossier page
Cluster route
Route the repository proof back into the thesis and founder question.
Cluster route / fireworks-tech-graph case study
Cluster route
Fit route
Turn cluster-level proof into a concrete working conversation.
Fit route / founder collaboration
Fit route
Commercial outcomes
Founding engineer fit
Open-source traction, shipped product surfaces, X field notes, and a founder-readable site.
Entry
Founding engineer collaboration page
Enter through the collaboration page for the founding-engineer path.
Fit route / founding engineer paths
Fit route
Proof
Repository dossiers and delivery surface
Use repositories and dossiers to prove delivery surface.
Proof route / repository dossiers
Proof route
Continuation
Thesis clusters and wider proof map
Then continue into thesis and the wider topic clusters.
Thesis route / topic clusters
Cluster route
Design partner sprint
The AI Devtool Sprint offer, proof clusters, and case-study style artifacts.
Entry
AI devtool sprint offer
Enter directly through the sprint offer.
Offer route / AI devtool sprint
Offer route
Proof
Workflow diagnosis in the archive
Return to the longform layer for workflow diagnosis.
Archive route / workflow diagnosis
Archive route
Continuation
Thesis clusters and wider proof map
Then continue into thesis and the wider topic clusters.
Thesis route / topic clusters
Cluster route
Developer-facing proof work
fireworks-tech-graph, project dossiers, technical diagrams, and launch narrative systems.
Entry
Flagship founder-readable case study
Enter through the flagship case for the strongest founder-readable proof.
Case page / fireworks-tech-graph
Case route
Proof
Live release cadence in the forum
Return to the forum for live cadence and release feedback.
Board route / release cadence
Board route
Continuation
Case back into the wider thesis map
Then route the case back into thesis and the wider topic map.
Cluster route / thesis map
Cluster route
Dossier atlas
Projects are not only listed by name. They are arranged by the workflow they replace.
Projects are not only listed by name. They are arranged by the workflow they replace.
01 / Python / 7.9k stars
Openfireworks-tech-graph
Hand-made one-off diagrams that collapse after a single meeting or slide deck.
This repo is market proof that the builder can turn vague technical structure into public artifacts developers want to share.
Topic continuation
Harness engineering for agentic AI products
What has to surround the model before an AI workflow becomes a product?
Start at AI Thesis, then continue through the thesis-side proof path.
Lead proof / AI Thesis
Cluster route
Topic continuation
Technical distribution for AI builders
How does technical work compound into trust, search traffic, and serious conversations?
Start at AI Thesis, then continue through the thesis-side proof path.
Lead proof / AI Thesis
Cluster route
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / fireworks-tech-graph
Dossier page
Case
Founder-readable flagship case study
This project also has a founder-readable case study to continue into.
Case page / fireworks-tech-graph
Case route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
02 / Python / 94 stars
Openfireworks-skill-memory
One giant memory dump that cannot distinguish a reusable pattern from random residue.
This is one of the clearest proofs on the site that AI memory should be structured, selective, and recoverable rather than merely larger.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / fireworks-skill-memory
Dossier page
Thesis
Wider thesis and proof graph
Return to the wider proof map.
Thesis route / proof graph
Thesis route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
03 / Python / 8 stars
Openfireworks-sessions-saver
Fragile long-running sessions where useful context disappears after pauses, crashes, or handoffs.
Interruption is not an edge case in real AI work. A team that can resume cleanly has a direct execution advantage.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / fireworks-sessions-saver
Dossier page
Cluster
Harness engineering for agentic AI products thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
04 / TypeScript / 2 stars
Openpaper-pal
PDF hoarding and one-pass reading that never turns into reusable understanding.
Serious AI work depends on reading systems, not just browsing papers. This project treats comprehension as production work.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / paper-pal
Dossier page
Thesis
Wider thesis and proof graph
Return to the wider proof map.
Thesis route / proof graph
Thesis route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Proof shelves
Viewed from the thesis side, which repositories provide execution proof?
Viewed from the thesis side, which repositories provide execution proof?
Cluster 01
Harness engineering for agentic AI products
Harness engineering for agentic AI products routes first into repository dossiers, then back out into the thesis, the forum, and collaboration.
Lead dossier
fireworks-tech-graph
2 execution nodes / start at the lead dossier
Dossier page / fireworks-tech-graph
Lead dossier
Lead proof
AI Thesis
3 onward routes from this cluster / continue through the thesis side
Cluster route / AI Thesis
Cluster route
Founder fit
Founder collaboration fit
Turn the cluster-level proof into a concrete working conversation.
Fit route / founder collaboration
Fit route
fireworks-tech-graph
7.9k starsHand-made one-off diagrams that collapse after a single meeting or slide deck.
This repo is market proof that the builder can turn vague technical structure into public artifacts developers want to share.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / fireworks-tech-graph
Dossier page
Case
Founder-readable flagship case study
This project also has a founder-readable case study to continue into.
Case page / fireworks-tech-graph
Case route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
fireworks-sessions-saver
8 starsFragile long-running sessions where useful context disappears after pauses, crashes, or handoffs.
Interruption is not an edge case in real AI work. A team that can resume cleanly has a direct execution advantage.
Dossier
Site-native README reading and proof ledger
Open the project page first for README rendering, site interpretation, and proof ledger.
Dossier page / fireworks-sessions-saver
Dossier page
Cluster
Harness engineering for agentic AI products thesis-side continuation
Place the repository back into the thesis and market question it supports.
Cluster route / thesis map
Cluster route
Fit
Founder collaboration fit
Let the repository signal flow into a concrete working conversation.
Fit route / founder collaboration
Fit route
Cluster 02
Long-running agents and resumable work
Long-running agents and resumable work routes first into repository dossiers, then back out into the thesis, the forum, and collaboration.
Lead dossier
Repository library
0 execution nodes / start at the lead dossier
Dossier page / repository library
Lead dossier
Lead proof
AI Thesis
3 onward routes from this cluster / continue through the thesis side
Cluster route / AI Thesis
Cluster route
Founder fit
Founder collaboration fit
Turn the cluster-level proof into a concrete working conversation.
Fit route / founder collaboration
Fit route
Cluster 03
Developer communication for AI infrastructure
Developer communication for AI infrastructure routes first into repository dossiers, then back out into the thesis, the forum, and collaboration.
Lead dossier
Repository library
0 execution nodes / start at the lead dossier
Dossier page / repository library
Lead dossier
Lead proof
fireworks-tech-graph case study
3 onward routes from this cluster / continue through the thesis side
Cluster route / fireworks-tech-graph case study
Cluster route
Founder fit
Founder collaboration fit
Turn the cluster-level proof into a concrete working conversation.
Fit route / founder collaboration
Fit route