makecindy
Featured
cindy
Cindy is an open-source AI agent that runs locally on your machine, integrates multiple AI harnesses and models, and provides memory, skills, and automation to perform real work in your projects and apps.
Danus
Danus is a system that uses a main agent (Claude Code) to coordinate a swarm of worker agents that prove mathematical claims, with a cold-start verifier gating acceptance and a content-addressed fact graph as the sole source of truth, ultimately producing human-readable reports or LaTeX papers.
Stars
106
7-day growth
No data
Forks
25
Open issues
0
License
Apache-2.0
Last updated
2026-07-22
AI repository intelligence
FR-AI / ANALYSISWhy it is worth attention
It enforces a strict separation of powers via role-gated tools, ensures correctness with a stateless verifier, scales proof search through parallel workers, and automatically verifies the final paper against the fact graph.
Who it is for
- Researchers in automated theorem proving and formal mathematics
- Mathematicians exploring large or complex proofs
- AI safety researchers studying aligned verification-driven agent architectures
- Developers building multi-agent reasoning systems with verifiable memory
Use cases
- Proving research-level mathematical theorems with long dependency chains
- Exploring multiple proof strategies in parallel and accumulating verified lemmas
- Generating publishable LaTeX papers directly from verified proof results
- Auditing and extending mathematical results with an immutable, dependency-tracked fact graph
Strengths
- Role-gated MCP server prevents unverified entries from entering the fact graph
- Stateless verifier ensures each submission is judged fresh with no memory contamination
- Content-addressed fact graph allows cascade revocation and deep dependency tracking (3,157 facts, 8,616 edges in one real run)
- Final paper is re-verified against the fact graph before delivery, building on per-fact verification
Considerations
- Requires bring-your-own API keys and an OpenAI-compatible codex backend, incurring token costs
- Users must explicitly define 'done' to avoid unbounded token spending on hard problems
- Write-paper output quality heavily depends on providing good example papers as exemplars
README quick start
Danus: Orchestrating Mathematical Reasoning Agents with Fact-Graph Memory
Danus orchestrates mathematical reasoning agents with fact-graph memory. A main agent (Claude Code) steers a swarm of autonomous codex workers that prove; a cold-start verifier is the sole authority on correctness: a result becomes real only once it passes. Verified results accumulate in a content-addressed fact graph — the system's only source of truth — and a strategy loop (a strong reasoning model) decomposes the problem and steers the swarm. When you have the answer, Danus renders it into a human report or a publishable LaTeX paper.
Danus builds on the worker–verifier core of our earlier system Rethlas (arXiv:2604.03789). The paper and the technical report tell the full story: the system, six research-level case studies it resolved, and what we learned along the way.
See ARCHITECTURE.md for the layered design and the map of every module.
How it works
The design follows a strict separation of powers: the main agent performs the global planning and coordination, the workers carry out the detailed proof search, the verifier is the sole authority on correctness, and the fact graph holds every verified result and is the system's only source of truth.
Each kind of agent carries out its role through its own skills and its own
role-gated set of tools, so the separation is enforced by construction, not by
prompts: the main agent has no fact_submit (the agent that steers the search
structurally cannot introduce unverified mathematics into the fact graph), and
the verifier writes nothing at all.
Every claim enters truth through one cycle:
A worker typically focuses on one claim at a time — a lemma, a counterexample, a toy example — rather than an entire proof. It repeatedly submits the claim with a supporting proof and revises it under the verifier's feedback until it passes, at which point the claim enters the fact graph as a fact, with the facts its proof depends on as its incoming edges. The verifier is stateless: a fresh instance judges each submission and retains nothing afterwards. Because each worker draws on only the facts it needs for its current claim and submits one fact at a time, the working
Description
Orchestrating Mathematical Reasoning Agents with Fact-Graph Memory
Related repositories
Similar projects matched by category, topics, and programming language.
7-e1even
Featured
learn-agent
A collection of notes on building coding agents, derived from the production development of the desktop agent Reina, with each mechanism simplified into a zero-dependency, single-file Node.js demo.
S40911120
Featured
recensa
Recensa is a self-hosted web viewer that indexes Claude Code session transcripts into a local SQLite database, enabling full-text search, replay, and audit of all past agent conversations without uploading data anywhere.