# Miucr

> Review code/diffs/PRs with the owned `miucr` CLI (miu-cr, a pure-Go AI code reviewer). Use when asked to review staged changes, a commit, a ref range, or a GitHub PR; to run/parse a gated review; compare reviewer quality with eval; to drive reviews over MCP; to run the serve webhook/poll daemon or GitHub Action; or to diagnose/upgrade miucr itself (over/under-approving, webhook/poll host issues, its docker-compose stack, binary version). Output is the stable `miucr.cli/v1` JSON envelope; parse it, don't grep prose. For an authored, opinionated PR review via gh, use vd:code-review.

- Skill: `vanducng/miucr` (Agent Skill)
- Install (CLI): `npx skillmds@latest add vanducng/miucr`
- Raw SKILL.md: https://api.skillmd.com/api/skills/vanducng/miucr/raw
- Safety review: WARNING
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: DevOps & Infra
- License: MIT
- Author: vanducng (https://skillmd.com/u/vanducng)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/vanducng/miucr

---


# miucr: owned AI code-review CLI (v0.65.0)

`miucr` (the **miu-cr** project) is a fast, **pure-Go** (`CGO_ENABLED=0`) AI code reviewer.
It keeps the correctness-critical parts **deterministic** (file selection, context assembly,
line-anchoring, severity gating, dedupe) and uses the LLM only for judgment (finding bugs,
proposing fixes). It runs five review ways:

- **Local review**: `miucr review` over a staged diff, a commit, or a ref range.
- **GitHub PR review**: `miucr review --pr` (dry-run by default; `--post` reacts 👀, upserts ONE summary issue comment, and posts inline comments as a PR review).
- **serve daemon**: HMAC webhook (default) and/or opt-in poll trigger; optional REST API + GitHub App auth.
- **MCP server**: `miucr mcp` exposes `review_run` / `review_get` over stdio to any agent host.
- **Evaluation**: `miucr eval` compares miu-cr and other reviewer commands against JSON expected findings.

**Review behavior worth knowing (design choices that prevent noise):**
- **One upserted summary, posted first.** `--post` writes ONE summary *issue comment*, edited in place on re-runs (never stacked), and acknowledges the PR with 👀. If no summary exists yet, it also creates a short `Review running` placeholder before the LLM starts; on later commits it keeps the prior completed summary visible until the new result is ready. The final summary replaces that same comment and is posted BEFORE the inline review so it anchors on top (overview → details). Inline findings are a separate PR review. `review_id` is NOT shown in the comment (it only resolves on the local store; it stays in the JSON envelope). On a **fatal review failure** after miucr's internal retries, `--post` upserts that SAME summary comment with a visible GitHub alert instead of failing silently: operational/provider/infrastructure failures (`agent.unavailable`, `provider.rate_limited`, `quota.exceeded`, `review.timeout`, `github.unavailable`, etc.) render as `[!WARNING]`, while unknown `internal.error` failures render as `[!CAUTION]`; a later successful run replaces it with findings.
- **The summary is a per-finding lifecycle ledger, not just the latest run.** Below a concise ≤5-bullet "What changed" summary, it renders two always-visible tracking tables - **⚠️ Open (N)** and **✅ Resolved (N)** - each finding tracked by its line-independent fingerprint across commits: a **Priority** column (P0-P4), status (`open` / `resolved` / `reopened`), the **origin commit** it was first raised on and the **resolved commit** it disappeared on (both linked), priority before→after for escalations, and first-seen / resolved timestamps. A clean review uses natural all-clear prose like `Review passed! No findings on the first review pass.` or `Review passed! 3 findings resolved. Good cleanup.` The deterministic note gets stronger as more findings are cleared and adapts to focused vs broad diffs. The footer is `Last reviewed commit` + review attempts + the miu-cr release. Lifecycle state is **storeless**: it lives in a hidden `<!-- miu-cr-ledger:<base64> -->` marker inside the comment (like the runs counter), so it survives ephemeral CI with no DB. A finding resolves only when it is absent AND its file is still in the diff (absence off-diff ≠ fix).
- **Inline comments persist; resolving the threads is left to you / your coding agent.** Each finding's inline comment is posted ONCE and deduped across re-runs via a hidden `<!-- miucr:fp=… -->` marker, so a re-review never re-posts or deletes it. miucr does NOT click "Resolve conversation": when a finding is fixed, GitHub auto-marks the thread *Outdated* and the summary ledger moves it to **✅ Resolved**, but the inline thread itself stays open for the developer/coding agent to resolve. A host config may opt into `thread_resolution_sync.mode: poll` to mirror manual GitHub "Resolve conversation" state into the summary (the Resolved row shows a clean `💬 conversation` marker linked to the discussion thread, distinct from a commit-resolved row's `open → resolved` SHA arrow); this never starts an LLM review and never feeds approval. So an agent acting on a review should read the inline threads, apply the fixes, reply on each handled inline thread with what changed and why, then resolve the thread.
- **Repeat-run stability is deterministic inputs + low-variance generation.** For the same repo/ref/config, file selection, context assembly, rules, anchoring, gating, fingerprints, and comment dedupe are deterministic. SDK-backed Anthropic/OpenAI calls use `temperature: 0`; exact model output can still vary, so PR posting is idempotent rather than duplicate-prone.
- **One-click suggestions and approvals are explicit write actions.** `--suggest` emits a native GitHub ` ```suggestion ` block ONLY for findings at or above the suggestion floor (`medium`) when the patch *deterministically* replaces the exact anchored line(s) and the model is certain of a grounded mechanical fix (a cited rule or an obvious best practice). It NEVER guesses an unverifiable value (a URL, path, route, ID, version, config key, API signature); such concerns become a verification-question in the rationale instead. `--patch-repair` (requires `--suggest`) runs one focused 2nd LLM pass to recover a repairable near-miss patch. `--approval clean|threshold` submits APPROVE only when the policy and safety gates pass; the first approval body is short (`LGTM`) and links to the code review summary unless `--approval-note none` is set. Approval is head-SHA scoped: a later clean push can be approved again bodyless, threshold re-approvals carry the threshold note, and the same commit is not approved twice. Permission/self-approval failures warn and degrade rather than failing the review.
- **`-o pretty`** is the human-readable local format; **`-o json`** is for agents; `-o sarif` for editors/CI.
- **Multi-provider profiles.** Add a named provider (e.g. z.ai/glm) with `kind`, `base_url`, `model`, `auth`, and either `auth_env` or `auth_command`; select with `--provider <name>`. Built-in kinds: `anthropic`, `openai` (ChatGPT-plan OAuth via `miucr login`). Transient GitHub/network errors auto-retry with backoff. Optionally cap a provider instance's usage with `[providers.<name>.quota]` (`dimension = tokens|requests`, `limit`, `window = <Go duration like 5h/24h>|monthly`); **uncapped by default**, fail-closed, over-quota → typed `quota.exceeded`.
- **Thinking on by default; deterministic fallback.** Capable models (Claude, gpt-5/o-series, codex, **z.ai GLM 4.5+**) review with **extended thinking/reasoning** (deeper analysis; temperature is omitted because thinking forces temp 1). Models without thinking (gpt-4o, plain glm-4 chat) sample at **temperature 0** for stable, reproducible findings. Both are config-exposed: `[review].thinking` (`auto|off|low|medium|high`, default auto) and `[review].temperature` (0-2, default 0). Set `MIUCR_TRACE_REASONING=true` to capture that reasoning into the review trace (`miucr trace <id>`).

## Output contract: `miucr.cli/v1` envelope (parse this)

Every command prints **one JSON object** on stdout (default `-o json`). Field order:

```json
{
  "ok": true,
  "api_version": "miucr.cli/v1",
  "kind": "review.result",
  "command": "review",
  "request_id": "req_...",
  "summary": { "findings": 2, "gate": "high" },
  "data": { "...": "command-specific" },
  "artifacts": [],
  "warnings": []
}
```

- `ok` (bool) is the branch point. `artifacts` and `warnings` are **always present** (`[]` when empty).
- `summary`, `data`, `page`, `stats` are present only when relevant (`omitempty`).
- Errors use the **same envelope** with `ok:false`, `kind:"error"`, and an `error` object:

```json
{ "ok": false, "api_version": "miucr.cli/v1", "kind": "error", "command": "review",
  "error": { "code": "review.gate_failed", "message": "<redacted>", "hint": "...",
             "retryable": false, "safe_to_retry": false } }
```

`kind` per command: `version`, `review.result`, `config.show`, `rules.check`, `rules.init`, `init.result`, `login.result`,
`whoami`, `logout`, `history.list`, `history.record`, `history.prune`, `eval.result`, `trace.show`, `error`
(REST: `review.accepted` / `review.result`). **Secrets never appear** in the envelope, logs, or on disk
(credential-named fields are scrubbed; finding `rationale`/`suggested_patch` prose is exempt).
The `--instruction`/`--conversation` flags and the `/miucr review <prompt>` comment trigger only add
**input** (a USER-turn context block); they never change the envelope or the finding JSON (still
`miucr.cli/v1`); parse the same shape.

### Exit codes (gate → exit mapping)

| Exit | Meaning |
| ---- | ------- |
| `0`  | Success: no finding reached `--gate`. |
| `1`  | Operational error: missing credentials, internal failure, store unavailable. |
| `2`  | **Gate failed** (a finding's severity ≥ `--gate`) **or** invalid invocation (bad gate, conflicting/zero modes, bad `--output`). |

Severities low to high: `info` < `low` < `medium` < `high` < `critical`. A `review.gate_failed`
error is emitted *after* the normal `review.result` envelope (the findings still print), then exit 2.

### Typed error codes (branch on `error.code`)

The day-1 provider/auth/timeout failures classify into a **stable taxonomy** (same code across all backends: anthropic/openai/codex), each with an actionable `hint` and a correct `retryable`:

| `error.code` | When | `retryable` | Hint |
| ------------ | ---- | ----------- | ---- |
| `agent.auth_failed` | bad/invalid API key (401/403, api-key backends) | `false` | `miucr login …` / set a valid key |
| `agent.auth_expired` | expired OAuth (401/403; codex incl. still-401-after-refresh) | `false` | `miucr login --provider openai` |
| `provider.rate_limited` | 429 | `true` | wait for the reset window and retry |
| `agent.unavailable` | 5xx / 529 | `true` | retry shortly |
| `review.timeout` | the review exceeded `--timeout` | `true` | raise `--timeout` (e.g. `1800s`) or narrow the diff |
| `review.canceled` | ctx canceled (Ctrl-C / SIGINT), exit `130` | `false` | - |
| `config.invalid` | malformed `config.toml` / bad enum or `auth` value / an `openai`-kind gateway profile with a key but no `base_url` (exit `2`; same code across review/history/serve) | `false` | fix the named field / set `base_url` for the gateway profile |
| `quota.exceeded` | the resolved provider's `[providers.<name>.quota]` is exhausted for the current window; exit `2`. On the serve host the PR is skipped+logged, not failed (a later push re-checks). A quota counter that can't be read/opened fails closed as the retryable `store.unavailable` instead (serve retries, not skips) | `false` | raise the provider quota `limit` or wait for the window to reset |
| `github.auth` | GitHub API hit `401`/`403` on fetch or publish (bad/missing `GITHUB_TOKEN` or insufficient scope) | `false` | check `GITHUB_TOKEN` / its repo scope |
| `github.pr_not_found` | PR fetch hit `404` (no such PR, or the token can't see it). Write-path 404s (deleted comment/reaction) keep the stage code (`github.upsert_summary_failed`, etc.) | `false` | check the PR exists and the token has access |
| `github.rate_limited` | GitHub API hit `429` (REST rate limit or abuse-detection). Envelope may include `error.details.retry_after_seconds` | `true` | wait for the reset window; the host requeues, in-process blip retries skip this code |
| `github.unavailable` | GitHub API hit `5xx` / a network error (DNS / refused / timeout / unexpected EOF from an idle HTTP/2 drop) | `true` | GitHub unavailable / unreachable, retry shortly |
| `github.pr_fetch_failed` | any other unclassified PR-fetch failure | `false` | - |
| `internal.error` | any unclassified failure (default; bare-wrapped) | `false` | - |

Unknown failures stay `internal.error` (never mislabeled as retryable). Classified messages are redacted: **no token fragment ever appears**.

The **codex** backend retries `429`/`502`/`503`/`504` (and a `response.failed` stream event) with bounded, jittered exponential backoff (≤3 attempts) like the SDK backends, honoring `Retry-After`/`resets_in_seconds` and aborting on cancel/timeout. A persistent rate limit returns `provider.rate_limited` with the usage-cap reset window in `error.details.resets_in_seconds` (or `retry_after_seconds`); branch on that to decide wait-vs-switch-provider.

## Install

```sh
curl -fsSL https://cr.miu.sh/install.sh | sh   # asset-aware latest
curl -fsSL https://cr.miu.sh/install.sh | sh -s -- v0.65.0   # pin
brew install vanducng/tap/miucr                                                     # Homebrew
go install github.com/vanducng/miu-cr/cmd/miucr@latest                              # Go 1.25+
```

Verify: `miucr version` → `{"ok":true,...,"data":{"version":"v0.65.0"}}`.
Config (optional) at `~/.config/miu/cr/config.toml`; state DB at `~/.config/miu/cr/state.db`.
Repo rules at `.miu/cr/rules/*.md` (**never a flat `.miucr/`**).

## Onboarding (`miucr init`)

`miucr init` is the fastest path to a working config. It walks a clean, sectioned
wizard (**provider → provider-aware auth → project rules**), then writes
`~/.config/miu/cr/config.toml` (dir `0700`, file `0600`, **deltas only**: the
chosen provider block, never the full built-in defaults) and ends on the literal
`miucr review --staged`.

```sh
miucr init                                  # interactive wizard (idempotent: Overwrite? y/N)
miucr init --non-interactive --provider anthropic --auth-env ANTHROPIC_API_KEY --yes
```

- **Provider-aware auth menu**: `openai` offers **browser login (OAuth, default)**
  (review on your ChatGPT/Codex plan, no API key) plus env-var or paste; it runs
  the same PKCE loopback flow as `miucr login` and caches the token in `oauth.json`
  (config records just `default_provider = "openai"`, no secret). `anthropic` offers
  env-var (default) or paste, **no OAuth** (Anthropic ToS). `custom` asks kind +
  base URL, then env-var or paste.
- **Default writes no secret**: only the env-var **name** (`auth_env`). A literal
  `auth_token` lands only on explicit paste + confirm (after a plaintext-on-disk warning).
- Flags: `--provider anthropic|openai|custom`, `--auth oauth|env|paste` (non-interactive
  selector), `--auth-env <NAME>`, `--base-url <gateway>`, `--no-rules`, `--force`,
  `--yes`, `--non-interactive`. `--auth oauth` is interactive-only (needs a browser);
  non-interactive errors `init.aborted` toward `miucr login`. Envelope `kind: init.result`
  (`data.auth_method` = `oauth|env|paste`, `data.next` = `miucr review --staged`); errors
  `init.aborted` / `config.write_failed`.
- `init` is **optional**: zero-config still works when a provider key is on the env.
  With no config **and** no key, `review` prints a soft one-line nudge to run `init`.

## Examples (copy-paste starters)

The repo ships an [`examples/`](https://github.com/vanducng/miu-cr/tree/main/examples)
tree: `rules/{go-api,typescript-node,python-data}.md`,
`github-action/code-review.yml` (fork-safe `pull_request_target`),
`workflows/miucr-review.yml` (the **dual-trigger** default: `pull_request` + a
`/miucr review <prompt>` comment trigger),
`mcp-setup/{claude-code,cursor,codex}` + `README-mcp.md`,
`review-local/{pre-commit,Makefile,agent-review.sh}` (review-your-own-changes
recipes), and `review-host/{Dockerfile,docker-compose.yml,config.example.yaml}`
(pure-Go `CGO_ENABLED=0` Alpine image + Postgres host stack for `miucr serve`).
Onboarding walkthrough lives at the docs
[Getting started](https://cr.miu.sh/onboarding/) page.

**Comment-triggered review (`/miucr review <prompt>`).** With the dual-trigger workflow
installed, a **write or admin collaborator** posts `/miucr review <prompt>` as a top-level PR
comment (e.g. `/miucr review focus on the auth changes`) to steer a re-review with free text.
The `issue_comment` event runs the trusted base-branch workflow with full secrets even on
fork PRs, so the workflow self-gates: the commenter must have **write|admin** (an
`actions/github-script` `repos.getCollaboratorPermissionLevel` check; `author_association`
alone is insufficient), the comment body is read via an env var (never inline-interpolated
into `run:`), the released binary is used (fork head code is never built with secrets), and
miucr adds a 👀 reaction to acknowledge an accepted command. Known limits (v1): only
top-level `issue_comment` is caught (not inline review comments or review summaries), and an
unchanged head SHA short-circuits (there is no `--force` on the comment path yet).

## Commands & exact flags

Global flags (all commands): `-o, --output json|pretty|sarif` (default `json`), `--timeout <dur>`
(default `30s`; `review` auto-bumps to `900s` and `eval` to `30m` unless `--timeout` is set explicitly).
`sarif` is **review-only**: it emits a SARIF 2.1.0 document (NOT the envelope) for
code-scanning/IDEs (`ruleId`=category, `level` from severity, repo-relative paths);
upload it with `github/codeql-action/upload-sarif`. `pretty` is a local reporter
(jumpable `file:line`, excerpt, patch; color on a TTY). `review --pr` also takes
`--filter-mode added|diff_context|file|nofilter` (default `diff_context`) controlling
which findings are inline-eligible; `file`/`nofilter` route off-diff findings to the
summary/SARIF/local output, never inline.

### `review`: needs **exactly one** mode

```sh
miucr review --staged                     # staged changes vs the index
miucr review --from main --to HEAD        # ref range (--from and --to required together)
miucr review --commit HEAD~1              # one commit vs its parent
miucr review --pr owner/repo#123          # a GitHub PR (dry-run by default)
miucr review --staged --instruction "focus on the auth changes"   # steer this one review
miucr review --pr owner/repo#123 --conversation                   # also read the prior PR thread
```

| Flag | Default | Notes |
| ---- | ------- | ----- |
| `--staged` | off | Review the **index** (what you're about to commit), not HEAD. |
| `--from` / `--to` | - | Range mode; **required together**. |
| `--commit <ref>` | - | Single commit vs first parent. |
| `--pr <url\|owner/repo#N>` | - | GitHub PR; `https://github.com/owner/repo/pull/N` or `owner/repo#N`. |
| `--gate none\|info\|low\|medium\|high\|critical` | `high` | Exit 2 when a finding reaches this severity. `none` never fails. |
| `--repo <dir>` | `.` | Repository directory. |
| `--include` / `--exclude` | - | Repeatable doublestar globs (path must match / drop). |
| `--ext go,ts,...` | - | Restrict to these file extensions. |
| `--expand <n>` | `5` | Context lines above/below each hunk (`0` disables). |
| `--token-budget <n>` | `0` | Approx token budget; over budget degrades context (`0` disables, so the default is no cap). |
| `--deep-context` | OFF | Heavier file context for large reviews: `--expand 20`, `--token-budget 0`, `--timeout 900s`, auto related-file hop depth, and root `AGENTS.md` / `CLAUDE.md` from the reviewed revision unless those flags are set. |
| `--context-hops <n>` | `0` | Override related-file context depth from changed files (`0` disables, max `5`); follows Go package imports/reverse imports and basic relative JS/TS/Python imports from the reviewed revision. Skipped on fork PRs. |
| `--provider anthropic\|openai\|<name>\|auto` | `auto` | LLM profile. |
| `--api-key` / `--base-url` / `--auth-token` / `--model` | - | Provider overrides; **never persisted**. |
| `--token <pat>` | - | GitHub PAT (overrides `GITHUB_TOKEN`/`GH_TOKEN`); required only for `--post`; never persisted. |
| `--post` / `--no-post` | `--no-post` (for `--pr`) | Publish vs dry-run; mutually exclusive (`flags.conflict`). |
| `--suggest` | OFF | Native one-click suggestions for proven fixes at or above `medium`: single-line replacements, contiguous range replacements, and wrap/guard/insert fixes on a QuotedCode-proven single-line anchor; requires `--post`; author-applied, never pushed. |
| `--patch-repair` | OFF | Conditional **2nd LLM pass** that recovers one-click suggestions the first pass *almost* produced: for each single-line finding `>= medium` whose `SuggestedPatch` was rejected for a *repairable* reason (empty / no-op / length mismatch - never a true anchor mismatch), one focused agent call asks for a minimal replacement of the verbatim anchored span, then **re-validates with the same exact-anchor gate**; emits the suggestion only if it now passes, else keeps the fenced hint. **Requires `--suggest`** (`config.invalid`, exit 2, otherwise); inert in dry-run (recovers only on `--post`). Bounded: per-review cap (default 5), highest-severity-first; one extra LLM call per repaired candidate. PR-path only, default OFF. |
| `--approval off\|clean\|threshold` | `off` | Submit `Event=APPROVE` by policy on `--pr --post`. `clean` requires zero findings; `threshold` allows findings at or below `--approval-max-priority`. Permission/self-approval failures warn and degrade - read `.data.pr.approve_reason` (e.g. `self_approve_forbidden` on your own PR) to tell "declined" from "not attempted" (`not_requested`). |
| `--approval-max-priority P0\|P1\|P2\|P3\|P4` | `P4` | Threshold approval ceiling. `P3` approves P3/P4 only; P0-P2 block approval. |
| `--approval-note none\|on_findings\|always` | `always` | Controls approval review body text. The default first approval body is short LGTM-style copy with a summary link; clean re-approvals after a later push can be bodyless. `none` suppresses it and `on_findings` only writes it when findings remain. |
| `--filter-mode added\|diff_context\|file\|nofilter` | `diff_context` | Inline-eligibility filter on `--pr`. `file`/`nofilter` route off-diff findings to summary/SARIF/local, never inline (GitHub 422s an off-diff comment). |
| `--min-severity none\|info\|low\|medium\|high\|critical` | none (no floor) | Minimum severity posted **inline** on `--pr`. Below-threshold findings still appear in the summary header counts + SARIF, never inline. An out-of-set value is rejected (`flags.invalid_min_severity`, exit 2). |
| `--walkthrough-diagram` | OFF | Opt in to a Mermaid change diagram in the summary (fenced ```mermaid block GitHub renders). Rides the same single review pass, no extra LLM call. Diagram quality varies; a malformed/omitted diagram degrades to a plain note. |
| `--mode review\|checks` | `review` | GitHub reporter on `--pr --post`. `review` posts inline comments + a summary. `checks` posts a GitHub CheckRun with annotations (survives force-push, works on fork PRs, can be a **required** check); conclusion maps from the gate (gate-clean→`success`, gate-hit→`failure`); needs `checks: write`. |
| `--format full\|minimal` | `full` | Review-comment presentation on `--pr`. `full` is the current output (summary section + severity/priority badges). `minimal` drops the `## Code Review Summary` section, **all** shields badges (summary chips **and** the per-finding inline `P3 · bug`), and the visible footer sub-line (the reviewed-head marker moves into a hidden `<!-- Reviewed commit … -->` token so the resolution-sync still finds it), keeping inline findings + the hidden upsert markers. Clean approvals still keep the lean summary comment so the approval can link to it. Render-only - does not change which findings surface. An out-of-set value is rejected (`flags.invalid_format`, exit 2). |
| `--walkthrough` / `--file-change-summary` | `true` / `false` | Toggle the two model-generated PR-overview blocks in the summary, independent of `--format`. `--walkthrough` (default ON) is the "What changed" bullets; `--file-change-summary` (default **OFF**) is the "Important Files Changed" per-file table - off by default to keep the summary lean. Render-only (the model still emits both; they're just dropped from the comment). They AND with `--format`, so `minimal` shows neither regardless. Also `[review].code_summary.{walkthrough,file_change_summary}` and host `review.code_summary`. |
| `--prompt-format xml\|markdown` | `xml` | Review **prompt** structure (orthogonal to `--format`, which is comment presentation). `xml` (the default) wraps untrusted payloads (diffs, new-content, project files, rules, conversation) in entity-escaped XML tags instead of `===`/`---` delimiters + fences, hardening against delimiter-collision / prompt-injection (a planted `</file>` or `=== File: ===` stays inert). `markdown` is the prior fenced form (byte-identical to ≤0.65), available as opt-out. Out-of-set value rejected (`flags.invalid_prompt_format`, exit 2). Also `[review].prompt_format`. |
| `--sarif-out <path>` | - | Also write a SARIF 2.1.0 report to `<path>` from the SAME single review run (in addition to `--output`/posting). Written only on success (atomic temp+rename); a failed run leaves no file. This is how the Action does single-pass SARIF, no second LLM call. |
| `--no-save` | off | Skip persisting this run to the local history store (every review is saved by default). |
| `--force` | off | On `--pr`, re-review even when the head SHA is unchanged since the last saved review. By default an unchanged head SHA short-circuits (`skipped_unchanged`, no LLM pass); a new commit always re-reviews. |
| `--instruction <text>` | - | Free-text steer for **this** review (e.g. `"focus on the auth changes"`). Injected into the **USER turn** as a fenced, context-only block; it never changes the finding rules, severity, category, or JSON schema, and rides the same single review pass (no extra LLM call). Trusted (developer-authored CLI flag); still UNTRUSTED when set from an `issue_comment` trigger on a fork PR. |
| `--conversation` | off | On `--pr`, fetch the prior PR conversation (miucr's summary + finding threads + developer replies) and inject it fenced/context-only as **UNTRUSTED** context (dropped on fork PRs). One extra GitHub **read** pass, no extra LLM call. |
| `-v, --verbose` / `-q, --quiet` | auto | Progress to **stderr** (stdout envelope unchanged). Auto-on when stderr is a TTY; `-v` forces on, `-q` forces off; mutually exclusive. Piped/CI stays silent. |
| `--trace` | off | Stream the live review trace (system prompt, diff, rules, prompts, response) as NDJSON to **stderr** (local-only, redacted; distinct from `--verbose`; stdout envelope unchanged). Inspect a saved review's trace with `miucr trace <id>`. |

**`review.result` data** (local and `--pr`):

```jsonc
"data": {
  "findings": [
    { "file": "internal/foo/bar.go", "line": 42, "end_line": 42,
      "title": "…optional short scannable summary…",   // omitted when the model emits none
      "rule": "go",                                     // optional: stem of the project rule that motivated this finding (omitted when none)
      "severity": "high", "category": "bug",
      "rationale": "…why this is a problem (may cite a convention the model can see, e.g. \"differs from mapWriteError\")…",
      "suggested_patch": "…optional minimal fix…",
      "quoted_code": "…verbatim source the finding anchors to…" }
  ],
  "stats": { "files_changed": 3, "files_reviewed": 2, "findings_total": 2,
             "findings_dropped": 1, "max_severity": "high", "gate": "high",
             "truncation_level": "full",            // full | hunks_only | filenames_only
             "rules_applied": 5, "rules_truncated": false },
  "review_id": "rev_...",  // additive: the saved history record id ("" only with --no-save; on an incremental skip it is the prior review id)
  // additive, only on a --pr run when an unchanged head SHA short-circuited (no
  // LLM pass). Dry-runs skip from the local history store; --post can skip from
  // the completed-publish marker in the summary comment. Both fields are omitted
  // on a normal review. On the skip path findings is [] and stats is {} (never null):
  "skipped_unchanged": true, "prior_review_id": "rev_prior",
  "pr": {  // only on --pr
    "owner": "owner", "repo": "repo", "number": 123, "head_sha": "deadbeef",
    "is_fork": false, "posted": false, "posted_inline": 0,
    "summary_action": "none",         // fate of the ONE upserted summary issue comment: none | created | edited | fork_fallback | failed
    "approve_action": "commented", "approve_reason": "not_requested", "suggestions_posted": 0,
    "patches_repaired": 0,            // additive, omitempty: findings whose rejected SuggestedPatch the --patch-repair 2nd pass recovered into a now-clean one-click suggestion (0/absent when --patch-repair is OFF)
    // additive, omitted when empty:
    "mode": "review",                 // review (default) | checks
    "check_run_id": 0, "check_conclusion": "",  // --mode checks only (success|failure)
    "fallback_annotations": 0 }       // >0 when a fork-PR 403 under Actions fell back to ::error:: workflow annotations (review did NOT hard-fail); summary_action then "fork_fallback"
}
```

`findings_dropped` = findings rejected by line-anchor drift (their quote no longer matches the reviewed
revision, kills position drift). `--post` keeps the summary and the inline findings in **separate
homes**: inline comments post as a PR **review** (body left empty, never a 422 on a no-inline run), and
the summary is **ONE issue comment that is UPSERTED**: miu-cr lists the PR's issue comments, finds the
one carrying the `<!-- miu-cr-review -->` marker, and **edits it in place**; if none exists it creates it.
So a re-run **updates the single summary** instead of stacking a review per commit. `summary_action` is
`created` (new summary issue comment), `edited` (upserted in place), `fork_fallback` (a fork PR lacked
comment-write scope, degraded, no hard fail), `failed` (summary upsert failed after inline posting), or
`none` (`--no-post`, `--mode checks`, or a skipped summary path). A same-commit `--post` re-run
short-circuits after a matching completed-publish marker; pass `--force` to review anyway.
per-comment `<!-- miucr:fp=... -->` line-free fingerprints prevent inline dupes across commits. The
summary body itself carries a **finding lifecycle ledger** - Open + Resolved tables with per-finding
status/origin+resolved commit/priority-before→after/timestamps - persisted storelessly in a hidden
`<!-- miu-cr-ledger:<base64> -->` marker so the open/resolved history survives across pushes (and
ephemeral CI) without a DB. The `miucr.cli/v1` JSON envelope is unchanged: the ledger is a
rendering/comment concern, not an envelope field.
A public-PR dry-run needs **no GitHub PAT** (LLM key still required); `--post` and private repos need a PAT with `repo` scope.

### `serve`: webhook daemon (default) + opt-in poll

```sh
WEBHOOK_SECRET=… GITHUB_TOKEN=… ANTHROPIC_API_KEY=… \
  miucr serve --addr :8080 --repos owner/repo,owner/other --gate high
```

| Flag | Default | Notes |
| ---- | ------- | ----- |
| `--addr` | `:8080` | Webhook listen address. |
| `--gate` | `high` | **Publish-severity only**: which findings get posted; never affects liveness/exit. |
| `--repos` | - | **Required** owner/repo allowlist (comma-separated); other repos are ignored. |
| `--poll` | off | Opt-in trigger: periodically ask GitHub which PRs need review. Webhook stays default. |
| `--poll-interval` | `1m0s` | Floor; effective = `max(this, X-Poll-Interval)`. |
| `--poll-source notifications\|pulls` | `notifications` | Candidate source. `pulls` = full coverage / cold-start-complete. |

Env: `WEBHOOK_SECRET` (required unless poll-only), `GITHUB_TOKEN`/`GH_TOKEN` (required unless `[github] mode=app`),
`ANTHROPIC_API_KEY` (or compatible). `MIUCR_LOG_LEVEL=debug` enables progress/tool-turn logs; `MIUCR_TRACE_LOG=true`
adds bounded debug trace payloads (`MIUCR_TRACE_LOG_MAX_BYTES`, default `4096`) that are redacted/truncated but may
include prompt/diff context. `MIUCR_TRACE_REASONING=true` captures the model's reasoning as a `reasoning` trace step
(Claude/GLM thinking verbatim; OpenAI a token count + `[hidden by provider]`) - off by default, redacted in storage,
needs `[review].thinking` on. Endpoints: `POST /webhook` (HMAC), `GET /healthz`. Each new head SHA = one full
LLM review; allowlist + per-head dedup are the only spend guards. serve inherits `--suggest` **OFF** and `review.approval.mode: off`.

In `serve --host` YAML, `thread_resolution_sync` is an object and defaults off. Enable it per repo when manual GitHub
conversation resolution should update the summary table between commits:

```yaml
review:
  thread_resolution_sync:
    mode: poll
    interval: 5m
```

**Opt-in REST API**: set `MIUCR_API_TOKEN` (env-only, no flag) to register `/v1`:

```sh
MIUCR_API_TOKEN=$(openssl rand -hex 32) WEBHOOK_SECRET=… GITHUB_TOKEN=… ANTHROPIC_API_KEY=… \
  miucr serve --addr :8080 --repos owner/repo
# Queue (202 + server-generated crypto/rand id):
curl -sS -X POST https://host/v1/reviews -H "Authorization: Bearer $MIUCR_API_TOKEN" \
  -H 'Content-Type: application/json' -d '{"owner":"acme","repo":"widgets","number":42}'
# Read back (whitelist: id,status,created_at,findings,stats, never the clone path):
curl -sS https://host/v1/reviews/<id> -H "Authorization: Bearer $MIUCR_API_TOKEN"
```

Status lifecycle: `pending` → `done`/`failed`. HTTP map: `400` bad body, `401` missing/wrong bearer
(empty token can never auth), `403` off-allowlist, `404` unknown id, `405` wrong method, `413` body > 64 KB.
**Single-operator**: one shared bearer = one trust boundary (not multi-tenant).

**GitHub App auth** (opt-in alternative to PAT): `[github] mode=app` in config (see below).

**Host mode**: `miucr serve --host` loads YAML from `MIUCR_CONFIG` or
`~/.config/miu/cr/host.yaml`, requires Postgres (`store.backend: postgres`;
`MIUCR_PG_DSN` preferred), and watches multiple repos from one daemon. Validate
without opening DB/secrets:

```sh
MIUCR_CONFIG=examples/review-host/config.example.yaml miucr serve --host --dry-run-config -o json
```

Dry-run emits envelope kind `serve.host_config` with a redacted config plus
summary counts (`repos`, `accounts`). Host YAML uses `providers`, `store`,
`github.accounts`, top-level `review`/`agent`, `host.review`, and `repos[]`;
layering is defaults -> `review`/`agent` -> `host.review` -> repo overrides.
PAT accounts support `auth_env`/`auth_file`/`auth_command`; App accounts support
App/installation ids from literal/env plus private keys from path/env/command.
Provider credentials also support `auth_env` and `auth_command`.

Repo prompts override the global prompt (`system_prompt` or
`system_prompt_file`, mutually exclusive). `repos[].rules` can point to markdown
files or a non-recursive directory of `*.md`; these are trusted host context and
cannot change the protected finding schema. Host review write policy lives at
the effective repo level: `post`, `force`, `suggest`, `patch_repair`,
`thread_resolution_sync`, and `approval`. First host mode does not push code.

Host poller state lives in Postgres: repos, PR sessions, queued jobs, attempts,
workspaces, poll cursors, and retention metadata. Startup applies versioned
schema migrations under an advisory lock. Claims use row locks for concurrency,
and `host.retention` prunes old jobs/attempts/sessions/workspace records/cursors.
The review-host compose example defaults `MIUCR_LOG_LEVEL=debug` for local
dogfood while leaving `MIUCR_TRACE_LOG=false`.

`review.pr_filter` layers top-level -> `host.review` -> `repos[].review`.
Draft PRs are skipped unless `include_drafts: true`. `default_action` defaults
to `include`; set `default_action: exclude` for allowlist repos. Rules are
evaluated in order and the last matching rule wins. Matchers inside one rule are
ANDed. Supported matchers: `authors`, `author_types` (`Bot`, `User`,
`Organization`), `author_associations`, `title_regexes`, `labels`,
`requested_reviewers`, `base_branches`, and `head_branches`. Use title regexes
for generated PRs that may be opened by a human token, such as `^chore\(deps\):`.
`comment_trigger_regexes` are used by the GitHub Action comment-trigger workflow,
not by `serve --host poll_source: pulls`.

```yaml
review:
  pr_filter:
    default_action: include
    include_drafts: false
    comment_trigger_regexes:
      - '(^|\s)(/miucr review\b|@vanducng\b)'
    rules:
      - action: exclude
        title_regexes: ['^chore\(deps\):']

repos:
  - name: critical-dbt
    slug: example-org/critical-dbt
    review:
      pr_filter:
        default_action: exclude
        rules:
          - action: include
            authors: ["vanducng"]
          - action: include
            requested_reviewers: ["vanducng"]
```

### `rules`: project review context

```sh
miucr rules init             # scaffold annotated .miu/cr/rules/example.md
miucr rules init --force     # overwrite
miucr rules check internal/foo/bar.go   # which loaded rules apply to a path (kind: rules.check)
```

Rule files are markdown with YAML frontmatter, then prose injected as **context only** (never gate):

```markdown
---
description: Project-specific review context for changes under cmd/.
globs:
  - "cmd/**/*.go"
  - "internal/**/*.go"
alwaysApply: false
context_files:
  - "AGENTS.md"
---
# Prose below the fence is injected as CONTEXT for the reviewer.
```

Three layers merged by **file stem** (later overrides earlier): built-in defaults (Trusted, lowest) →
user `~/.config/miu/cr/rules/*.md` (Trusted) → repo `.miu/cr/rules/*.md` (**Untrusted**, highest).
A file with **no `---` fence is skipped** (never always-applied). Untrusted repo rules are fenced
context-only, byte-capped, and **dropped on fork PRs**; the finding-JSON schema stays in the cached
system prompt so injected prose can't redefine it. `rules check` data lists each applicable rule with
`provenance`, `stem`, `globs`/`always_apply`, `trusted`, plus skipped `body_only` files.

**Rule grounding.** A finding may carry the `rule` stem of the project rule that motivated it. The
wire layer validates the stem against the rules actually loaded this review (a hallucinated stem is
dropped) and renders it as `(per <stem>)` on the inline comment + summary overflow. A repo rule
(`.miu/cr/rules/*.md`) additionally links to its file, repo-relative at the head SHA; user and
built-in rules are cited as text only (no link, a user-rule home path never leaks).

### `history`: browse saved reviews

Every review auto-saves a **full record** (findings + stats + per-turn transcript + raw prompt/response)
to the local store; `--no-save` opts out per run. Records are local only (`~/.config/miu/cr/state.db`,
gitignored); **no tokens are ever stored**.

```sh
miucr history                                 # list recent reviews, newest first (kind: history.list)
miucr history --repo owner/repo               # filter by repo (PR) or repo dir (local)
miucr history --pr owner/repo#7               # filter to one PR
miucr history --since 7d                       # 7d / 24h / 2026-06-01
miucr history --limit 50                        # cap rows (default 20; 0 = no limit)
miucr history show <id>                         # one full record (kind: history.record; 404 → history.not_found)
miucr history show <id> -o pretty --raw         # pretty, with raw prompt/response inline
miucr history prune --keep 200 --yes            # keep newest N (kind: history.prune; reports deleted count)
miucr history prune --older-than 30d --yes      # delete records older than a span
```

`prune` needs at least one of `--keep`/`--older-than` plus `--yes` (destructive). An optional
`[history] max_records = N` auto-prunes oldest on save. List rows carry
`{id, created_at, target, mode, findings, max_severity, status}`; `show` data adds
`provider`, `model`, `head_sha`, `findings`, `stats`, `transcript`, `raw_prompt`, `raw_response`.
Errors: `history.unavailable`, `history.not_found`, `history.prune_policy_required`,
`history.prune_confirm_required`, `history.bad_pr`, `history.bad_time`.

### `eval`: compare reviewer quality against expected findings

`miucr eval` runs one or more reviewer commands over a JSON suite and scores them by
file+line overlap against expected findings. Use synthetic/public fixtures only.
For unlabeled public-PR smoke suites, read unmatched findings as review output to
ins

…(truncated)