doc-bridge

npm: @agentskit/doc-bridge · CLI: ak-docs · Landing: agentskit-io.github.io/doc-bridge

Topics: ai-agents · documentation · developer-experience · mcp · llms-txt · typescript

Compatibility: node >=22 · TypeScript 5.8+ · pnpm, npm, or yarn consumers

Turn your docs into executable handoffs for coding agents.

doc-bridge reads your repo docs, ownership map, and human documentation site, then gives humans and agents the same evidence-linked starting point:

• where to start reading
• which files/packages it may edit
• which checks prove the change
• which human docs explain the feature

It is not a wiki or hosted RAG. The core works without any LLM or API key; the documentation portal dogfoods AgentsKit Chat as an optional surface over that deterministic layer.

Built for agent-scale repositories

Doc Bridge turns large repository structure and documentation into compact, evidence-linked context that humans and coding agents can query instead of repeatedly traversing the full repository.

Historical context-payload estimate

An earlier anonymized dogfooding cycle estimated up to 99% less serialized context payload. This is a historical payload estimate—not a guarantee of provider-token savings, answer quality, or semantic correctness. It is not the same measure as provider-token usage.

Historical controlled A/B signal

The published controlled study with 96 anonymized executions reported a directional operational signal of:

• 18.46% fewer paired provider token-equivalent units across 46 token-complete pairs;
• 39.75 seconds lower P95 latency;
• 87.5% operationally completed executions vs. 75.0% with repository-only context.

These are historical, separately defined measures: the 99% figure is an estimated context-payload reduction, while the 18.46% figure uses provider-token-equivalent data from 46 paired observations. The bounded adjudicator recorded zero adjudicator-success outcomes in both arms; because that adjudicator is mechanical and does not independently judge semantic correctness, this result is directional and inconclusive. A newer local pilot is intentionally not promoted here while its semantic evaluation and publication review remain incomplete. See the full methodology and anonymized data.

Why teams use it

Agents are powerful, but most repo docs are written for humans. The result is familiar: the agent guesses ownership, edits the sibling package, runs the wrong test, or ignores the human guide that already explained the rule.

doc-bridge works in both directions:

Direction	What it does	Command
Human docs → agents	Turns Fumadocs, Docusaurus, markdown, and ownership docs into AgentHandoff	ak-docs index · ak-docs query --agent
Agent memory → docs	Reads .agent-memory/** and .cursor/rules/*.mdc, classifies what should become project docs, and drafts a human-reviewed promotion	ak-docs memory ingest · classify · promote --pr

The handoff is a routing contract:

{
  "startHere": "docs/for-agents/packages/auth.md",
  "editRoots": ["packages/auth"],
  "checks": ["pnpm --filter @demo/auth test"],
  "humanDoc": "/docs/guides/auth"
}

Workflow runs may carry the same optional correlation envelope used by the AgentsKit runtime and Chat protocol. operationId is the cross-repository identity; runId, sessionId, turnId, actionId, and traceId retain local meaning. It is bounded metadata only and must not contain prompts, secrets, or document content.

That contract works from the terminal, MCP, CI, and optional RAG/chat.

Documentation quality and reconciliation

Discovery is only the first step. ak-docs audit documentation compares declared documentation and ownership with the observed project graph and reports evidence-backed findings for missing coverage, stale relations, structured contradictions, exact duplicates, missing examples, and incomplete maintenance metadata.

Natural-language correctness, unnecessary prose, and semantic redundancy remain explicitly not-analyzed until a configured agent or human review evaluates them. Proposed changes stay reviewable and human-approved.

Example finding (anonymized)
  CONTRADICTION · high confidence
  Documentation declaration differs from the observed project relation
  Evidence: 4 source files + 1 documentation declaration
  Action: review ownership and update the canonical document

60-second proof

This README owns the one-command proof; the Getting started guide owns the complete repository setup and first-index workflow.

npm i -D @agentskit/doc-bridge
npx ak-docs demo --text

No config, no docs to read first. Output shows before/after, a real handoff, gate red→green, and the MCP snippet:

After (handoff.resolve / query --agent)
  ✓ target:  auth (packages/auth)
  ✓ start:   docs/for-agents/packages/auth.md
  ✓ edit:    packages/auth
  ✓ checks:  pnpm --filter @demo/auth test · pnpm --filter @demo/auth lint
  ✓ human guide: /docs/guides/auth

Gate: red → green

Verify the real handoff path

This checked example runs the bundled demo through the public CLI. The repository's scripts/check-readme-standard.mjs gate compares this block byte-for-byte with the executable fixture; .github/workflows/ci.yml runs that gate and its executable tests on every pull request.

The study figures below are not recomputed by the README gate; their protocol, privacy checks, and limitations are documented in the linked study artifacts.

import { execFileSync } from 'node:child_process'

execFileSync(process.execPath, ['bin/ak-docs.js', 'demo', '--text'], {
  stdio: 'inherit',
})

node examples/verify-handoff.mjs

For the complete repository setup, follow the Getting started guide.

Using Cline? Follow the deterministic llms-install.md setup. It runs the pinned MCP server through pnpm dlx without adding Doc Bridge to your repository dependencies.

What ships

See the surface map for a visual overview of the CLI, MCP, CI, and adapter surfaces.

Surface	Use it for	Command / artifact
CLI	Inspect ownership, search docs, run gates, ask local questions	ak-docs query, search, ask, doctor, gate
MCP server	Let Cursor, Claude Code, Codex-style agents resolve handoffs before editing	ak-docs mcp, handoff.resolve
GitHub Action / CI	Fail stale indexes and configured documentation gates on PRs	AgentsKit-io/doc-bridge@ee756a13c006c597445c31e2643c1e8cece715d7
Documentation conformance	Check the stable ecosystem standard with auditable evidence	ak-docs conformance run documentation-standard-v1 --text
Documentation audit	Measure documentation quality and compare docs with the observed project graph	ak-docs audit documentation --json
Doc adapters	Link human docs to agent docs	fumadocs, docusaurus, vitepress, starlight, nextra, plain-markdown
Monorepo routing	Discover workspaces and checks	pnpm-monorepo, nx
Memory pipeline	Turn agent notes into reviewable documentation drafts	memory ingest, classify, promote --pr
Optional RAG/chat	Ground chat in the same handoff-first index	@agentskit/rag, @agentskit/ink, ak-docs chat

See docs/getting-started.md, docs/mcp.md, and docs/examples.md.

Cursor plugin

This repository also contains a Cursor plugin that pairs the Doc Bridge MCP server (read-only except for the explicit proposal tool) with a handoff skill. It resolves startHere, readBeforeEditing, editRoots, and checks before Cursor edits a routed repository. The plugin does not request credentials or write project files through MCP.

GitHub Copilot plugin

The root Agent Plugins manifest exposes the same portable handoff skill and MCP server to GitHub Copilot CLI. Copilot discovers skills/ and .mcp.json from the standard plugin layout, so the integration stays source-owned instead of copying prompts into another repository.

copilot plugin install AgentsKit-io/doc-bridge

Portable Agent Skill

skills/doc-bridge-handoff packages the same fail-closed routing contract in the open Agent Skills layout for OpenClaw-compatible clients, Hermes Agent, Pi, Cursor, and other runtimes that can execute a local skill script. The skill prefers the read-only MCP tool and falls back to a pinned, zero-credential CLI resolver. It never edits files, runs returned checks, or grants authority outside editRoots.

Install the published skill from ClawHub:

clawhub install doc-bridge-handoff

Pi users can install the same source-owned skill through the npm package:

pi install npm:@agentskit/doc-bridge

Claude Desktop MCP Bundle

Doc Bridge can be packaged as a local MCP Bundle for Claude Desktop. The bundle advertises 14 MCP tools, exercises 8 of them in its smoke test, and marks every tool except docbridge.proposals read-only. It asks the user to select the repository's doc-bridge.config.json; that file defines the project boundary Doc Bridge may read.

From a clean checkout:

pnpm install --frozen-lockfile
pnpm mcpb:pack

The command builds Doc Bridge, creates a production-only staging directory, validates the MCPB manifest, packs the extension, checks its file inventory, and writes the local artifact under .mcpb-output/. Generated bundles and staging directories are intentionally excluded from Git.

The current packaged compatibility declaration is macOS-only; this project does not claim bundle support for other operating systems until the exact artifact passes an independent installation test there.

Why this exists

Pattern	Gap
Wiki + RAG	Explains; weak on where to act and proof docs match code
AGENTS.md alone	Great static rules; no ownership index, gates, or human bridge
Context7-class tools	Library docs for the model; not your monorepo routing

doc-bridge ships AgentHandoff JSON:

{
  "type": "agent-handoff",
  "startHere": "docs/for-agents/packages/auth.md",
  "editRoots": ["packages/auth"],
  "checks": ["pnpm --filter @demo/auth test"],
  "humanDoc": "/docs/guides/auth",
  "bridge": { "humanDoc": "linked" }
}

When a human guide is missing, handoffs surface it as a feature:

{
  "bridge": {
    "humanDoc": "missing",
    "action": "ak-docs bootstrap agent-docs"
  },
  "notes": ["Human guide missing for billing. Run: ak-docs bootstrap agent-docs"]
}

Four loops (with real commands)

Loop	Command	What you see
Act	ak-docs query package auth --agent	editRoots, checks, startHere
Bridge	ak-docs bootstrap agent-docs	Draft agent docs from human site; bridge.humanDoc in handoff
Learn	ak-docs memory classify → promote	HITL draft for agent corpus
Explain	ak-docs ask "auth is broken in staging"	Ownership match + handoff preview + next commands

ak-docs ask "who owns schemas"
# Best match: ownership os-core
# Handoff preview
#   start:  docs/for-agents/packages/os-core.md
#   edit:   packages/os-core
#   checks: pnpm --filter os-core lint · pnpm --filter os-core test

Coverage your team checks daily

The following is illustrative output from the command, not a measurement of this repository. Run the command locally or in CI for current values.

ak-docs doctor --text
ak-docs doctor --badge          # shields.io markdown for README
ak-docs index --watch           # keep index fresh while editing docs

Score: 82/100 (B)
  Agent docs:      8/10 (80% handoff-ready)
  Human guides:    6/10 (60% bridged)
  Gates:           3/3 passing

Next actions
  → ak-docs bootstrap agent-docs
  → ak-docs query package billing --agent

Agent uses it alone

1. MCP auto-wire: ak-docs mcp install --cursor
2. Skill/rule: paste docs/skills/doc-bridge.md into Cursor rules — agents call handoff.resolve before editing packages/*
3. Handoff is the next step: startHere, checks, and bridge are in the JSON/MCP response

CI as first-class citizen

Reuse the bundled GitHub Action on every PR:

Follow the canonical Gate and CI guide, which includes the complete workflow and pins the published Action release.

The Action installs the exact configured package (or the workspace package when dogfooding this repository), then verifies the committed index and configured gates without silently rebuilding them. It rejects non-exact package versions. See the Marketplace guide.

The guide is pinned to the published stable Action release v1.7.45; the checked-in package version is 1.11.1.

Coverage is repository-specific. Run ak-docs doctor --badge locally to emit current handoff and human-bridge badges, or pnpm coverage:badge in CI; this README intentionally avoids publishing a stale static percentage.

Or locally, as two steps rather than one, because they answer different questions:

ak-docs index     # after changing docs or config — then review and commit the result
ak-docs gate run  # verifies the committed index, which is what CI verifies

Chaining them (index && gate run) cannot report a stale index: it gates an artifact written a second earlier against a rebuild of the same tree. The gate's value is that the committed index and the repository agree, so run it the way CI does, against what is committed. It fails with Index is stale. Run: ak-docs index — the same check, and the same annotation, as in CI.

Product surface

Core — always (no LLM)

Surface	Purpose
Demo	ak-docs demo — bundled fixture, no setup
Doctor	Coverage score, missing humanDoc/agent doc, next actions
Index	DocBridgeIndex + contentHash + llms.txt + capabilities
CLI	query / search / list / ask / gate / bench / memory / bootstrap
MCP	handoff.resolve, doc.search, doc.get, gate.status, …
Gates	Freshness, human-link validation, optional OKF style
Adapters	pnpm-monorepo, nx, fumadocs, docusaurus, vitepress, starlight, nextra, plain-markdown

Optional AgentsKit peers

npm i -D @agentskit/rag @agentskit/ink @agentskit/adapters @agentskit/memory react
ak-docs rag ingest && ak-docs chat

See docs/chat-and-rag.md.

AgentsKit ecosystem

Who uses it (public)

Designed for and dogfooded on open AgentsKit surfaces:

Surface	Link
for-agents	agentskit.io/docs/for-agents
Registry	registry.agentskit.io
Playbook	playbook.agentskit.io
AgentsKit Chat	documentation · source
Code Review	repository-native CLI
This repo	CI green · ak-docs gate run on every PR

Playbook pattern: docs/playbook/doc-bridge-pattern.md — export with ak-docs playbook pattern --text

Configuration examples

Profile	Example
Solo markdown	examples/minimal-plain-markdown.config.ts
pnpm monorepo	examples/pnpm-monorepo.config.ts
Nx monorepo	examples/nx-monorepo.config.ts
Demo monorepo	examples/demo-monorepo/
Fumadocs + chat	examples/fumadocs-with-chat.config.ts
VitePress	examples/vitepress-only.config.ts
Astro Starlight	examples/starlight-only.config.ts
Nextra	examples/nextra-only.config.ts

Contract: docs/spec/config-v1.md · CLI: docs/spec/cli.md · MCP: docs/mcp.md · Skill: docs/skills/doc-bridge.md · Pattern: docs/playbook/doc-bridge-pattern.md · Recipes: docs/recipes/index-pipeline.md

Learn loop — memory → draft PR

ak-docs memory ingest
ak-docs memory classify
ak-docs memory promote --pr --dry-run   # preview gh commands
ak-docs memory promote --pr              # opens draft PR via gh

Status

Published npm package: v1.7.45 stable. The working-tree package version is 1.8.0 and is not published yet. The Action example below intentionally pins the latest published stable release; the release workflow updates the package and Action version together when a new release is published.

The published package provides portable, fail-closed handoffs through the CLI, MCP server, CI action, and packaged skill; deterministic Documentation Standard v1 conformance; verified release provenance; and documentation-quality audit tooling.

pnpm install && pnpm build && pnpm test
pnpm smoke:ollama    # optional — skips if Ollama/peers unavailable

Landing: https://doc-bridge.agentskit.io/

Privacy Policy

The local MCP server reads only the project selected through doc-bridge.config.json. It does not require an API key, send project data to AgentsKit, collect telemetry, or write project files through its MCP tools. See the complete Privacy Policy for accessed paths, use, storage, sharing, retention, optional integrations, and contact information.

Contributing

Issues and PRs are welcome. Start here:

To improve the evidence base, reproduce the anonymized study, add a language or framework analyzer, contribute a documentation-quality rule, or add a fixture for a real contradiction or stale relation.

Need	Doc
Local setup, tests, release flow	CONTRIBUTING.md
Governance and maintainer responsibilities	GOVERNANCE.md
Vulnerability reports	SECURITY.md
Community standards	CODE_OF_CONDUCT.md
Release history	CHANGELOG.md
Product positioning	docs/POSITIONING.md

License

MIT