review-bench
Skill2 starsChanged 19 days agoMITView on GitHub
Benchmarks qa-review on one pull request, and any other code reviewer the user names against it: runs each reviewer headless several times, builds a verified findings ledger from qa-review warm-up runs, grades every run against the frozen ledger, and prints a score table of mean quality, run-to-run sd, criticals found and noise tiers. Use when the user asks to benchmark, score, rank or compare code reviewers, review commands, models or effort levels against qa-review, or to measure a reviewer's recall on a pull request.
In this repository16 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- ask-teamSkill
- assemble-metaSkill
- brainstormSkill
- build-issueSkill
- cut-milestonesSkill
- frictionSkill
- milestone-closeSkill
- qa-reviewSkill
- scope-commitSkill
- scope-issueSkill
- scope-mrSkill
- scope-specSkill
- setup-dream-teamSkill
- ship-itSkill
- triage-issueSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
16 rules. About 0 are common, 3 are distinctive, and 13 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
13
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 16
Read <skill-dir>/METHODOLOGY.md and the usage header of <skill-dir>/run.sh before step 1.
- Very specific: line 18
Run every run.sh call with Bash run_in_background.
- Very specific: line 19
Run run.sh calls on one <out> one at a time: wait for each to finish before the next.
- Very specific: line 21
Before step 3 and before reporting, check every <out>/<run>/marks.txt exists and is non-empty.
- Very specific: line 29
Read baseRefOid and headRefOid from gh pr view <n> -R <owner>/<repo> --json baseRefOid,headRefOid.
- Very specific: line 31
Use git merge-base <baseRefOid> <headRefOid> as <base-sha>, <headRefOid> as <head-sha>.
- Very specific: line 32
Run git checkout --detach <head-sha> in the clone.
- Very specific: line 40
Write <out>/qa-high.txt as qa-high:lenses model=<model-id> /dream-team:qa-review {base}.
- Very specific: line 52
Delete <out>/qa-high-r{1..3}/marks.txt and <out>/qa-max-r{1,2}/marks.txt.
- Distinctive: line 59
No reviewer named → skip to step 6.
- Very specific: line 62
Write each reviewer as a <PREFIX>[:lenses] model=<model-id> [effort=<e>] [shown=<name>] <prompt> line in a file under <out>.
- Very specific: line 64
Write a version in a prefix as 4-6, never 4.6.
- Very specific: line 66
Write the prompt so the reviewer reviews <base-sha>..<head-sha>.
- Very specific: line 79
After each run.sh call from step 3 on:
- Distinctive: line 82
Check each against the code: drop one that is refuted, pre-existing, already a row, or a duplicate of another candidate.
- Distinctive: line 86
Only minor candidates survive → bump nothing.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
review-bench
<skill-dir> below is this skill's base directory, as the harness prints it.
- Read
<skill-dir>/METHODOLOGY.mdand the usage header of<skill-dir>/run.shbefore step 1. run.shruns, grades and scores; this skill covers the calls it cannot make.- Run every
run.shcall with Bashrun_in_background. - Run
run.shcalls on one<out>one at a time: wait for each to finish before the next; usePfor parallel runs, and parallel grades on a frozen ledger, inside one call. - Before step 3 and before reporting, check every
<out>/<run>/marks.txtexists and is non-empty; if one is not, rerun the samerun.shcommand, which regrades only the ungraded runs.
1. Pick the pull request
- Pick a pull request in a public repo whose diff holds real defects and fits one review run.
- Clone the repo outside the output directory and fetch the head:
git fetch origin pull/<n>/head. - Read
baseRefOidandheadRefOidfromgh pr view <n> -R <owner>/<repo> --json baseRefOid,headRefOid. - Use
git merge-base <baseRefOid> <headRefOid>as<base-sha>,<headRefOid>as<head-sha>. - Run
git checkout --detach <head-sha>in the clone; step 5 checks candidates against it. - Pick an output directory
<out>outside the clone; every step below uses the same one.
2. Warm up the ledger
- Name a qa-review reviewer file
<out>/qa-<effort>.txtand its prefixqa-<effort>, so its run directories readqa-high-r1,qa-max-r1. <model-id>below is a full model ID, such asclaude-opus-4-8, never an alias such asopus.- Write
<out>/qa-high.txtasqa-high:lenses model=<model-id> /dream-team:qa-review {base}. - Write
<out>/qa-max.txtasqa-max:lenses model=<model-id> /dream-team:qa-review {base}. - Run
N=3 EFFORT=high <skill-dir>/run.sh <clone> <base-sha> <head-sha> <out>/qa-high.txt <out>. - Then run
N=2 EFFORT=max <skill-dir>/run.sh <clone> <base-sha> <head-sha> <out>/qa-max.txt <out>. - These five runs are the warm-up; no stop rule ends it earlier or later.
- A warm-up run with
early-exitand a non-zeroexit=instats.txt→ delete<out>/<run>and rerun its command before step 3. - The score tables these calls print grade against a growing ledger; they are not results.
3. Freeze the ledger
- Set
"frozen": truein<out>/ledger.json; leaveversionat 1. - Delete
<out>/qa-high-r{1..3}/marks.txtand<out>/qa-max-r{1,2}/marks.txt. - Rerun the step 2
qa-high.txtcommand to regrade all five; the warm-up runs stay scored runs. - The frozen ledger lacks a
criticalor animportantrow → drop this pull request and go back to step 1 with another and a new<out>.
4. Run each reviewer the user named
- No reviewer named → skip to step 6; the table holds qa-review alone.
- qa-review at another effort →
<out>/qa-<effort>.txtper step 2, run withEFFORT=<effort>. - A reviewer named later → reuse the same
<out>and run steps 4 to 6. - Write each reviewer as a
<PREFIX>[:lenses] model=<model-id> [effort=<e>] [shown=<name>] <prompt>line in a file under<out>; prefix is ASCII letters, digits and-, unique per reviewer. - Write a version in a prefix as
4-6, never4.6. :lensesonly for a reviewer whose subagents each return a report.- Write the prompt so the reviewer reviews
<base-sha>..<head-sha>;{base}alone is a base ref only to qa-review; another reviewer may read a bare SHA as one commit. - An external reviewer whose review runs on another model, such as
/vllm-reviewon pi → addshown=<that model>, such asshown=qwen3-coder;model=still sets the Claude session's model. - A reviewer differing only in model or effort needs its own prefix, such as
qa-opus-4-6, and its ownmodel=oreffort=field; onerun.shcall can hold all such reviewers. - A reviewer whose prompt and plugin have not run on this pull request → run
N=1, check itsreport.mdcovers the pull request's files, then rerun withN=3; the rerun skips the first run. - A reviewer whose prompt and plugin already ran on this pull request → run
N=3in one call. run.shgrades each new run against the frozen ledger and prints the score table.
5. Promote candidates
After each run.sh call from step 3 on:
- List
grep -h ^Candidate <out>/*/marks.txt. - Check each against the code: drop one that is refuted, pre-existing, already a row, or a duplicate of another candidate.
- A non-minor candidate survives → add every surviving candidate as a row per METHODOLOGY
"Ledger file", and bump
versionby one. - Only minor candidates survive → bump nothing; they wait for the next bump.
- Any other row change (severity,
plausible) also bumpsversionby one. - On a bump, read each earlier run's
report.mdorlenses/*.mdagainst the new rows only, per METHODOLOGY "Grading a run"; add the IDs each report or lens earns to its own scored line; existing marks stand. - Rescore:
cat <out>/*/marks.txt > <out>/marks.txt, thenpython3 <skill-dir>/score.py <out>/ledger.json <out>/marks.txt.
6. Read and report
- Reviewers in one tier are not ranked; their order is noise.
- With 3 runs a reviewer's sd is often 3 to 7 points; a mean gap inside that is not a result.
- Compare scores only within one ledger version.
early-exitin a run'sstats.txtscores 0; name the run.
- **Ledger:** <score.py summary line>; pending candidates: <n or none>
- **Table:**
<score.py table as a markdown table, its columns and values verbatim>
- **Reading:** <tiers in words; gaps within sd called noise; rejected claims and $/run beside
quality; a multi-lens reviewer's rejected claims are raw lens claims>
- **Early exits:** <runs with their reviewer and model, or none>Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

