Start here, then follow the path that matches what you are doing. New users should read QUICKSTART.md first; the full README.md is the complete reference.
- QUICKSTART.md — install and first calls in under five minutes
- getting-started.md — guided first-session walkthrough
- why-zigars.md — what problem zigars solves vs. plain shell
- agent-clients.md — wiring Codex, Claude, Gemini CLI, Hermes
- codex.md — Codex-specific setup
- cli.md — thin JSON CLI, exit codes, follow-up commands
- tools.md — tool reference and evidence-tier overview
- tool-index.generated.md — generated catalog of every tool
- skills.md — the 19 shipped
@zigars/skills - backends.md — optional backend setup (ZLS, ZLint, zwanzig, zflame, diff-folded)
- troubleshooting.md — common errors and fixes
- evidence-tiers.md — command/LSP/parser/heuristic labels
- trust.md — process trust posture and the trust manifest
- determinism.md — what "deterministic" means here
- maturity.md — feature-area maturity rubric
- architecture.md — layer boundaries and handler contracts
- agent-workflows.md — agent workflow tool design
- testing.md — test suites, fuzzing, coverage
- perf.md — performance and benchmarking
- ci-artifacts.md — CI ingestion tooling
- dogfooding.md — developing zigars with zigars
- release.md — release checklist and gates
- release-evidence.md — assembling the evidence package
- distribution.md — npm shim, MCPB, GitHub release channels
- security-model.md — workspace sandbox and path policy
- security-audit.md — vulnerability handling