{"id":44388673,"url":"https://github.com/decapodlabs/decapod","last_synced_at":"2026-05-28T06:11:17.574Z","repository":{"id":337856911,"uuid":"1155300084","full_name":"DecapodLabs/decapod","owner":"DecapodLabs","description":"Decapod is a Rust-built, local-first governance runtime for AI agents: a repo-native knowledge graph and state store, enforced workflows, proof-gated changes, and safe coordination across any loop or model.","archived":false,"fork":false,"pushed_at":"2026-02-15T09:50:20.000Z","size":12797,"stargazers_count":3,"open_issues_count":0,"forks_count":2,"subscribers_count":1,"default_branch":"master","last_synced_at":"2026-02-15T12:12:59.262Z","etag":null,"topics":["agents","ai","ai-agents","control-plane","kernel","orchestration","rust"],"latest_commit_sha":null,"homepage":"https://ko-fi.com/decapodlabs","language":"Rust","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/DecapodLabs.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":"CONTRIBUTING.md","funding":".github/FUNDING.yml","license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":"SECURITY.md","support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":"AGENTS.md","dco":null,"cla":null},"funding":{"github":["alexhraber"],"patreon":null,"open_collective":null,"ko_fi":"decapodlabs","tidelift":null,"community_bridge":null,"liberapay":null,"issuehunt":null,"lfx_crowdfunding":null,"polar":null,"buy_me_a_coffee":null,"thanks_dev":null,"custom":null}},"created_at":"2026-02-11T11:00:58.000Z","updated_at":"2026-02-15T09:50:20.000Z","dependencies_parsed_at":"2026-02-17T09:00:52.313Z","dependency_job_id":null,"html_url":"https://github.com/DecapodLabs/decapod","commit_stats":null,"previous_names":["decapodlabs/decapod"],"tags_count":103,"template":false,"template_full_name":null,"purl":"pkg:github/DecapodLabs/decapod","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/DecapodLabs%2Fdecapod","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/DecapodLabs%2Fdecapod/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/DecapodLabs%2Fdecapod/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/DecapodLabs%2Fdecapod/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/DecapodLabs","download_url":"https://codeload.github.com/DecapodLabs/decapod/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/DecapodLabs%2Fdecapod/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29538810,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-17T08:11:05.436Z","status":"ssl_error","status_checked_at":"2026-02-17T08:09:38.860Z","response_time":100,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.6:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agents","ai","ai-agents","control-plane","kernel","orchestration","rust"],"created_at":"2026-02-12T01:55:41.562Z","updated_at":"2026-05-28T06:11:17.569Z","avatar_url":"https://github.com/DecapodLabs.png","language":"Rust","funding_links":["https://github.com/sponsors/alexhraber","https://ko-fi.com/decapodlabs"],"categories":[],"sub_categories":[],"readme":"\u003cp align=\"center\"\u003e🦀\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003ccode\u003ecargo install decapod \u0026\u0026 decapod init --proof\u003c/code\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003cstrong\u003eDecapod\u003c/strong\u003e\u003cbr /\u003e\n  Daemonless, local-first governance kernel behind AI coding agents.\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  Agents call Decapod on demand to turn intent into context, then context into explicit specifications before inference,\u003cbr /\u003e\n  enforce boundaries, and deliver proof-backed completion across concurrent multi-agent work.\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"https://github.com/DecapodLabs/decapod/actions\"\u003e\u003cimg alt=\"CI\" src=\"https://github.com/DecapodLabs/decapod/actions/workflows/ci.yml/badge.svg\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://crates.io/crates/decapod\"\u003e\u003cimg alt=\"crates.io\" src=\"https://img.shields.io/crates/v/decapod.svg\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://github.com/DecapodLabs/decapod/blob/master/LICENSE\"\u003e\u003cimg alt=\"License: MIT\" src=\"https://img.shields.io/badge/license-MIT-blue.svg\"\u003e\u003c/a\u003e\n\u003c/p\u003e\n\nCanonical Contract: `assets/constitution.json` section `core/DECAPOD`\n\n---\n\n## Documentation\n\nDecapod provides comprehensive documentation for both human operators and AI agents.\n\n- **[Human Documentation (mdBook)](docs/book/src/introduction.md)**: Conceptual overview, workflows, adoption guide, and reference.\n- **[Agent Orientation Corpus](docs/agent/api-index.md)**: API-awareness layer for agents, including command contracts and payload examples.\n- **[Universal Agent Contract (AGENTS.md)](AGENTS.md)**: The machine-readable entrypoint for all agents operating in this repo.\n\n## Quick Start\n\n```bash\ncargo install decapod\ndecapod init --proof\n```\n\n`decapod init --proof` creates `.decapod/`, a local folder your agent uses to remember intent, rules, context, specs, and proof.\n\nYour **conversational** workflow does not change. You keep working through your agent; Decapod gives the agent the missing control plane. Intent is captured, scope is bounded, context is shaped, protected areas are respected, work is isolated, and completion is proven against the project’s rules and the Decapod constitution.\n\n---\n\n## How it works\n\nAI coding agents often lose the plot: they forget intent, pull too much context, skip dependencies, and touch protected files. Decapod gives them a repo-native governance layer that makes intent explicit, boundaries enforceable, context deliberate, and completion provable.\n\n### The Loop\n\n```mermaid\nflowchart TD\n    UserIn[\"User\"] --\u003e|\"intent\"| AgentPre[\"Agent (Pre)\"]\n    AgentPre --\u003e|\"governed request\"| Model[\"Model\"]\n    Model --\u003e|\"response\"| AgentPost[\"Agent (Post)\"]\n    AgentPost --\u003e|\"verified result\"| UserOut[\"User\"]\n\n    AgentPre -.-\u003e|\"ping for context\"| UserIn\n\n    AgentPre -. \"optional governance path\" .-\u003e DecapodPre[\"Decapod (Pre)\"]\n    DecapodPre -. \"intent, context, gates\" .-\u003e AgentPre\n\n    AgentPost -. \"optional proof path\" .-\u003e DecapodPost[\"Decapod (Post)\"]\n    DecapodPost -. \"boundaries, checks, proof\" .-\u003e AgentPost\n    DecapodPost -. \"needs more context\" .-\u003e AgentPre\n\n    style UserIn fill:#ff6b9d,stroke:#c44569,color:#fff\n    style UserOut fill:#ff6b9d,stroke:#c44569,color:#fff\n    style AgentPre fill:#a855f7,stroke:#7c3aed,color:#fff\n    style AgentPost fill:#a855f7,stroke:#7c3aed,color:#fff\n    style Model fill:#06b6d4,stroke:#0891b2,color:#fff\n    style DecapodPre fill:#fbbf24,stroke:#f59e0b,color:#000\n    style DecapodPost fill:#fbbf24,stroke:#f59e0b,color:#000\n```\n\n**Agent ↔ User pings** — The 1st agent (governance) and 2nd agent (proof) can ping the user for additional context when intent is unclear or verification needs human input.\n\nDecapod is called by the agent at governance boundaries. Before inference, the agent may branch into Decapod to shape intent, context, and gates. After inference, the agent may branch into Decapod when the work needs boundary checks, verification, proof, or another governed pass.\n\nEach Decapod call may recurse until the work is shaped, bounded, and provable. Decapod is not the agent and not the model; it is the governance kernel the agent calls whenever work needs control.\n\nDecapod is called before:\n\n- **Acting** — clarify intent and generate specs\n- **Inference** — resolve focused context capsules\n- **Touching Code** — enforce boundaries and protected paths\n- **Completing** — produce verification and proof\n\n---\n\n## Capabilities\n\n1. **Clarifies intent** — Converts vague requests into explicit, versioned specifications.\n2. **Bounds context** — Resolves only the minimal relevant code/docs for the task.\n3. **Enforces boundaries** — Safeguards protected branches and sensitive modules.\n4. **Governs adaptation** — Manages feedback-driven instruction changes through explicit review.\n5. **Requires proof** — Gates completion on deterministic verification artifacts.\n\n---\n\n## The substrate\n\nDecapod preserves what agent workbenches lose: reusable, repo-native knowledge that survives the session.\n\n```text\n.decapod/\n  generated/\n    specs/         # Human-visible intent and architecture specs\n    context/       # Deterministic context capsules\n    artifacts/     # Verification output and proof provenance\n  data/            # Durable repo-native state (DBs, events, todos)\n  config.toml      # Project shape and agent-facing configuration\n  OVERRIDE.md      # Local rules that override embedded defaults\n```\n\nEvery run leaves operational evidence. The generated files are the human-visible proof surface: inspect them locally, review them in PRs, and use them to re-establish state across different agents like Claude, Codex, Gemini, Cursor, and Kilo.\n\n---\n\n## The constitution\n\nDecapod ships with an embedded engineering constitution: over 100 declarative documents covering architecture, security, performance, and testing.\n\nEverything an engineering org usually keeps in tribal memory or review culture becomes executable guidance. Your agent does not guess; it reads the constitution, cites claim IDs, follows gates, and produces proof.\n\n---\n\n## Guarantees\n\n- **Daemonless** — Runs on demand like `git` or `grep`.\n- **Repo-native** — All state lives in your repository.\n- **Provider-agnostic** — Works across agent workbenches.\n- **Proof-gated** — Completion requires passed verification gates.\n- **Boundary-aware** — Enforces protected paths and branch isolation.\n\n---\n\n## Contributing\n\n```bash\ngit clone https://github.com/DecapodLabs/decapod\ncd decapod\ncargo build \u0026\u0026 cargo test\n```\n\n- [CONTRIBUTING.md](CONTRIBUTING.md)\n- [SECURITY.md](SECURITY.md)\n- [Issues](https://github.com/DecapodLabs/decapod/issues)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdecapodlabs%2Fdecapod","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdecapodlabs%2Fdecapod","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdecapodlabs%2Fdecapod/lists"}