"Why did Claude touch src/session.rs?
That wasn't in the ticket."
— Every engineering team, every week
Stores structured metadata in .git/.h5i/ — no new infrastructure, no lock-in, works with any git remote.
Prompt, model, agent ID, and test results linked to every commit.
Files read vs. edited, uncertainty heatmap, blind-edit detection.
Version-controlled OBSERVE / THINK / ACT trace that survives session resets.
Enforce AI commit rules and generate audit reports across date ranges.
A new agent starts with the branch goal, milestones, open TODOs, and the last useful decisions.
OBSERVE, THINK, ACT, NOTE, branch, merge, and checkpoint entries form a Git-native context DAG.
When the branch reaches review, h5i renders the same context into a sticky GitHub PR comment.
$ h5i codex prelude [h5i] Context workspace active goal: ship OAuth login safely milestones: provider integration done; token refresh in progress TODO: integration test for failover path Last decisions: THINK Redis survives process restarts ACT switched session store to Redis $ h5i share pr post --style review # same context, rendered for reviewers
Version-controlled OBSERVE / THINK / ACT trace stored in refs/h5i/context. Snapshots on every commit. Branches like git.
Log each reasoning step as OBSERVE, THINK, or ACT — persists across session resets.
BM25 over traces + git co-change. Replaces exploratory Read chains with ranked candidates.
Agents burn tokens every session re-deriving what they already figured out last time. h5i capture claim records each conclusion with its evidence pinned as a Merkle hash over the files it depends on — stays live until any evidence blob changes, then auto-invalidates.
Evidence = sha256 over (path, blob_oid) pairs at HEAD. Git tells you when they're stale — no TTL, no guessing.
Live claims render as ## Known facts at session start. Agent treats them as pre-verified — skips re-reading.
── A/B · 5 trials per arm ───────────────── metric No claims With claims Δ ──────────────── ───────── ─────────── ───── Read tool calls 6.2 ± 0.4 4.0 ± 0 −35% Cache-read tok. 793,425 511,824 −35% Assistant turns 25.6 ± 3.8 16.6 ± 3.1 −35% Est. session cost ~$4.35 ~$2.13 −51% Task fidelity 5/5 5/5 ✓ ✔ All 5 seeded trials read exactly the 4 relevant HTTP files.
Define TOML rules — require audit flag, minimum prompt length, model allowlist, test coverage threshold.
Checks staged files against policy before every commit. Blocks on errors, warns on violations.
AI contribution ratio, audit coverage, policy violations, and top-risk files over any date range.
$ h5i audit policy check ✔ model declared (claude-sonnet-4-6) ✔ audit flag present ✗ test metrics missing [error] ⚠ prompt length < 20 chars [warn] 1 error · 1 warning — commit blocked $ h5i audit compliance --since 2026-04-01 ── Compliance Report ────────────────────── AI-generated: 74% of commits (30 days) Audit coverage: 91% Policy errors: 3 warnings: 8 Top risk file: src/billing/token.rs
12 deterministic rules — credential leaks, CI tampering, scope creep.
Per-line authorship with agent name, model, and test badge.
Non-destructive restore to any past commit. WIP auto-backed up.
Scans the reasoning trace for prompt-injection patterns.
Snapshots Claude's memory at every commit, diffs across versions.
Timeline, Integrity, Intent Graph, Memory, Sessions at localhost:7150.