Skip to content

INDEX.md

Latest commit

 

History

History
267 lines (253 loc) · 34.2 KB

INDEX.md

File metadata and controls

267 lines (253 loc) · 34.2 KB
Raw

Docs Index

A single entry point for human readers and AI agents walking the docs/ tree.

Start here. A human evaluating one change wants quickstart.md — one review end to end on a committed sample, including which build provides which commands. A coding agent wants ../AGENTS.md and agents/README.md. The repository README.md is the landing page that routes to both.

Three constructed host-review examples show shell permissions, a remote MCP declaration and workflow token permissions, with exact Git histories and captured output from released 1.1.0. The asset bundle includes a short demo and transcript; publication and reader observations remain separate, pending work.

Concepts

  • overview.md — one-page summary for developers, reviewers, and AI agents
  • concepts.md — tool-use readiness in depth (the seven dimensions)
  • mental-model.md — the 5-minute model: one engine, one verdict, every artifact a projection; who reads what and what agents cannot cheat
  • report-v1-consolidation-rc.md — proposal to regroup the report's top level by reader and freeze v1.0 (no behavior change yet)
  • hosted-plane-design.md — design boundary for a future optional hosted product over local attestations/registry rows; the gate stays local
  • mcp-governance.md — the two MCP surfaces (agent tool exports vs coding-agent host grants) and the SHIP-HOST-BOUNDARY-* checks
  • host-boundary-support.md — exact static Codex/Claude Code/Cursor/VS Code/GitHub support and scope matrix
  • mcp-server.md — optional local MCP server mode (mcp-serve): read-only shipgate.check, shipgate.preflight, shipgate.explain, and shipgate.capabilities
  • organization.md — local-first organizational governance: policy-pack pins, exception hygiene, host-grant drift, attestations, and registry reporting
  • category.md — what an "agent release gate" is, in product terms
  • glossary.md — category vocabulary
  • ai-search-summary.md — human-readable summary for AI search and coding agents
  • design-partners.md — early design partner criteria and contact path
  • design-partner-verifier-pilot.md — runbook for the design-partner cohort: the two routes under test, the denominators, what counts as reviewer-understood first value, and the pre-registered continue/narrow/stop rule
  • design-partner-pilot-results.md — the public aggregate ledger for that experiment: denominators, dated enrollment shortfall, reproduced blockers, and the standing decision
  • architecture.md — codebase layout for new contributors
  • decisions.md — current accepted-decision index, superseded v0.1 defaults and explicit implementation obligations before the v1.0 freeze
  • engineering/ai-coding-workflow-verifier.md — canonical engineering guide and roadmap for making Agents Shipgate the deterministic verifier inside AI coding workflows
  • engineering/insufficient-evidence-cold-start.md — proposed design for getting first-adoption repos out of a standing insufficient_evidence verdict on every turn
  • engineering/host-authenticated-approval-receipts.md — proposed design for in-session approval receipts a host attests, and the record of why the unsigned version was rejected
  • agent-native-merge-contract.md — the agent-native protocol map: the eight merge contracts, each mapped to the artifact that implements it
  • product-hardening-gap-closure.md — closure map for the root dogfood gate, governance case catalog, policy-pack tests, trace contract, and runtime-inventory boundary
  • agent-workflow-evidence.md — local Agent Workflow Evidence and AgentTraceEvent contract for replayable verifier scenarios
  • capability-standard.md — stable static capability lock/diff standard for external integrations and research
  • capability-payload.md — the frozen shipgate.capability_payload/v1 payload shared by the exported capability delta and the committed capability state, and v2, which adds each tool's definition digests; one payload, two views, one subject per row
  • capability-delta-attestation.md — the capability delta published as a standalone in-toto attestation any consumer can verify without running Agents Shipgate; predicate type, subject binding, and the reference verifier
  • passed-verdict-contract.md — evidence-backed static meaning of passed, semantic gap routing, and 0.15 migration
  • governance-benchmark.md — stable research benchmark for evaluating agent governance behavior
  • manifest-v0.1.md — manifest schema in prose form
  • trust-model.md — what the scanner does and doesn't do
  • baseline.md — baseline workflow
  • framework-adapter-checklist.md — checklist for adding static framework adapters
  • mcp-registration-idioms.md — the 30-server survey behind the built-in MCP registration-idiom registry: which shapes ship, what the reader excludes, and why an export still wins
  • distribution-surfaces.md — every surface this engine is published through, what each one claims, and the test that proves it; the registry a new surface has to be added to
  • determinism-boundary.md — generated coverage matrix: what each input can establish per declaration shape, the extraction-confidence ceiling it reaches, and what that ceiling means for a verdict

Reference

Examples

Workflows

  • quickstart.md — the human entry path: one review end to end on a committed sample — which build you get, what the change added, why the top result matters, what was not established, and who owns the next action; then the two adoption routes, advisory CI, and the second PR
  • faq.md — common questions, AI-search-friendly
  • integrations.md — CI/CD integration recipes (GitHub Actions, GitLab CI, CircleCI, Jenkins snippet)
  • troubleshooting.md — error messages → fixes
  • distribution.md — release process and SBOM/signature verification
  • release-runbook.md — cutting a tag: mandatory rehearsal, the two-job publication transaction, provenance bindings, and the recovery path when PyPI succeeds but finalization fails
  • release-evidence-policy-decision.md — the approved release evidence bar: the 38-case pre_1_0 policy for 0.x tags, the 80-case beta policy from 1.0 on, and the promotion path between them (decided 2026-08-29, #341; amended 2026-09-04, #520)
  • release-evidence-policy-decision.md § Amendment 2 — the unqualified preview channel: the admissibility finding, the five conditions it holds under, and what would have made it inadmissible (decided 2026-09-02, #491)
  • changelog/1.2.0.md — the full reviewed prose for every 1.2.0 change; CHANGELOG.md carries the one-line-per-change release note
  • changelog/1.1.0.md — the full reviewed prose for every 1.1.0 change; CHANGELOG.md carries the one-line-per-change release note
  • changelog/1.0.0.md — the full reviewed prose for every 1.0.0 change made after the 0.16.0 section was cut; CHANGELOG.md carries the one-line-per-change release note
  • changelog/0.16.0.md — the full reviewed prose for every 0.16.0 change; CHANGELOG.md carries the one-line-per-change release note
  • decisions.md — architectural decisions

For agents