A skill your agent uses when configuring or changing a repository's tracked branches, layered review policy, ignored tracked paths, retrieval limits, summary depth, or non-secret task-board metadata.

Apache-2.0Auto-check passed

Install Configure Review

skills CLI
$ npx skills add hashgraph-online/awesome-codex-plugins --skill configure-review -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install hashgraph-online/awesome-codex-plugins configure-review --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/hashgraph-online/awesome-codex-plugins.git skills-src && mkdir -p .claude/skills && cp -r skills-src/plugins/mimfort/rag_for_git/plugin/skills/configure-review .claude/skills/configure-review && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
configure-review
GitHub stars
1.3k
Token cost
~3.8k tokens
SKILL.md length
1,860 words
Files
1
Skills in repo
716
Repo updated
First seen
Licence
Apache-2.0

At a glance

A skill your agent uses when configuring or changing a repository's tracked branches, layered review policy, ignored tracked paths, retrieval limits, summary depth, or non-secret task-board metadata.

  • Works in 5 steps: Resolve the canonical lowercase… → After the target is selected, verify the… → Scan only tracked Python files → …
  • Changing a repositorys tracked branches
  • SKILL.md covers Scope, Safe YAML preflight, Pipeline and Repository branches, plus 4 more sections
  • Calls git

What it does

Configure Review is an agent skill from hashgraph-online/awesome-codex-plugins. Use when configuring or changing a repository's tracked branches, layered review policy, ignored tracked paths, retrieval limits, summary depth, or non-secret task-board metadata.

Its SKILL.md is about 3.8k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

The repository describes itself as: A curated list of awesome OpenAI Codex / ChatGPT plugins, skills, and resources. The 1 Codex Marketplace. See live plugins at: https://hol.org/plugins/best-codex-plugins. The licence is Apache-2.0.

When your agent uses it

  • Changing a repositorys tracked branches
  • Layered review policy
  • Ignored tracked paths
  • Retrieval limits

Example prompts

  • “/configure-review”

Requirements

  • Python 3

Workflow steps

5 steps, taken from the first numbered list in SKILL.md.

  1. Resolve the canonical lowercase repository id and the target branch. Present these targets in
  2. After the target is selected, verify the repository with git rev-parse --git-dir, run the Safe
  3. Scan only tracked Python files
  4. Measure churn with git log --since="6 months ago" --name-only --pretty=format: -- '*.py'.
  5. Propose depth and ignore changes. Ask the user about every candidate for paths.ignore and

What it can do on your machine

Read from SKILL.md and the folder at commit 3e1456a. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • git

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use git, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Configure Review loads about 3.8k tokens when it runs. Until then it costs about 49 tokens; SKILL.md has 1,860 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~49
When it runs · the whole SKILL.md, loaded when a task matches
~3.8k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from hashgraph-online/awesome-codex-plugins at commit 3e1456a, republished under its Apache-2.0 licence (© hashgraph-online). 1,860 words, ~3,758 tokens.

Download SKILL.mdSave it as .claude/skills/configure-review/SKILL.md (or your agent's skills folder).
name
configure-review
description
Use when configuring or changing a repository's tracked branches, layered review policy, ignored tracked paths, retrieval limits, summary depth, or non-secret task-board metadata.

Configure Review

Always answer the user in Russian. Update only the requested policy values; Never clobber foreign keys such as categories. This skill is standalone: no reviewer MCP, database, or board connection is required for the baseline.

Scope

Manage summary_cluster_depth, summary_cluster_depth_overrides, summary_topk_threshold, summary_paths.ignore, paths.ignore, context_limits, and the optional task_board block, including its generic sync_filter. An empty task_board: disables the board for this repository. Never read, request, display, or write credential values in either policy target.

Tracked branches are separate from review policy. Manage repository.primary_branch and repository.index_branches only in the home per-repo target. The committed .review.yml cannot own repository, because branch selection must be available before a committed ref can be read.

Untracked .venv, node_modules, __pycache__, dist, and build are gitignored and never enter the index. Do not use a filesystem walk to find them.

Safe YAML preflight

Inspect each selected policy or home YAML file with a local boolean-only process. Run this preflight before any tool call that can return file contents. Return only safe/blocked, never matching lines, values, or exception text. Do not use Read or Grep to perform this preflight. The process must reject non-regular or symlinked files, malformed or non-mapping YAML, and credential-like keys at any depth. Also reject duplicate mapping keys, including duplicate repository keys and duplicate branch fields. Reject anchors, aliases, and merge keys. Reject these cases before reading or mutating the file in model context. Only after a safe result may a content-returning tool read the file for a line-oriented edit.

For a home target, derive the canonical home config root lexically even when it does not exist. Check every existing parent path component through the destination without following symlinks; reject symlinks and non-directories. Missing destinations, including a missing home config root, are allowed only when the nearest existing parent is a real directory and the normalized destination remains inside the canonical home config root. After creating any missing directories, run the path preflight again. Also re-check immediately before writing so a changed path never inherits an earlier safe result.

Pipeline

  1. Resolve the canonical lowercase repository id and the target branch. Present these targets in this order and ask the user to select one:
    • Recommended/default: home:repos/<owner>/<name>.yml, stored at $XDG_CONFIG_HOME/rag-reviewer/repos/<owner>/<name>.yml (or ~/.config/rag-reviewer/... when XDG_CONFIG_HOME is unset). It needs no commit and is not visible to the team.
    • Team-visible: committed .review.yml at the selected target ref. It is committed and visible to the team; read it from that ref, never from an uncommitted worktree file. For a nested id such as group/service, use home:repos/group/service.yml. A home policy is owned by the OS account running reviewer: on a shared service account it can affect that account's workloads, so use committed policy for team-owned settings.
  2. After the target is selected, verify the repository with git rev-parse --git-dir, run the Safe YAML preflight, and only then read the selected file, preserving unrelated keys and comments. Do not inspect or copy credentials.
  3. Scan only tracked Python files:
    bash
    git -C <path> ls-tree -r --name-only <branch> | grep '\.py$'
    Count directory prefixes at depths 1–3. This is not a filesystem walk.
  4. Measure churn with git log --since="6 months ago" --name-only --pretty=format: -- '*.py'. If history is too short or unavailable, say so and recommend from structure alone.
  5. Propose depth and ignore changes. Ask the user about every candidate for paths.ignore and never write it silently. Assemble a draft that preserves the selected file's unrelated keys/comments, then request final confirmation before writing it. Follow the exact rebuild map below; suggest but do NOT run a follow-up skill. When branch and policy changes share a run, assemble both drafts first, show both paths and diffs, and request one final confirmation before either write.

Repository branches

Handle branches before policy analysis whenever the user asks to inspect or change tracked branches.

  1. Resolve the local repository without network calls:
    • git rev-parse --show-toplevel gives the git root;
    • git remote get-url origin gives the canonical SSH/HTTPS remote candidate;
    • normalize it to lowercase <owner/name> with the same SSH/HTTPS forms accepted by reviewer;
    • if origin is absent or unrecognized, ask for <owner/name> explicitly. Network git commands are forbidden.
  2. Run reviewer config show --repo <owner/name> --json and show the effective primary branch, ordered index branches, and source. A policy/VCS diagnostic error does not erase the returned branch section; a malformed home config is a blocking error and must not fall back silently.
  3. Ask for repository.primary_branch, then ask for the complete ordered unique repository.index_branches. The primary must be present in the index list. Reject empty names, duplicates, and a primary outside the list.
  4. The destination is always $XDG_CONFIG_HOME/rag-reviewer/repos/<owner>/<name>.yml (or the equivalent ~/.config/rag-reviewer/... path when XDG is unset). Never write repository to committed .review.yml, even when committed policy is selected for other keys. If policy and branches change together, treat them as two targets in one preview.
  5. Run the Safe YAML preflight on the destination before reading it. Stop on every blocked result. Build a line-oriented patch: if repository is absent, append the canonical block; if it exists, replace only primary_branch and index_branches. Preserve all top-level keys and unknown repository subkeys. Preserve comments, line endings, and surrounding YAML style. Never serialize the complete file with yaml.safe_dump.
  6. Show the destination, source, old and new branch values, and the exact patch. Request one final confirmation before any branch or policy write. A rejection leaves every target unchanged.
  7. After writing, run reviewer config show --repo <owner/name> --json again and require the exact primary/index/source expected from the home per-repo layer. Report a mismatch as an error.

If newly added index branches are not indexed, suggest rag-reviewer:sync-codebase once per new branch, but do not run it. A primary change to an already indexed branch needs no rebuild. Removing a branch stops reviewer from selecting it but does not delete its old base index automatically. Branch changes never trigger subsystem-summary work.

Rebuild guidance

  • Changed paths.ignore → suggest rag-reviewer:sync-codebase.
  • Changed summary_cluster_depth → suggest rag-reviewer:summarize-subsystems.
  • Changed summary_cluster_depth_overrides → suggest rag-reviewer:summarize-subsystems.
  • Changed summary_paths.ignore → this key is part of the summary layout_token; suggest rag-reviewer:summarize-subsystems and warn it forces a full rebuild of every subsystem summary (same layout_token invalidation as summary_cluster_depth/ summary_cluster_depth_overrides).
  • Changed summary_topk_threshold → no rebuild needed.
  • Changed context_limits → no rebuild needed.
  • Changed task_board.sync_filter → suggest rag-reviewer:sync-tasks for a full unlimited run (limit=null); do NOT run it automatically.

Generic board metadata

Ask whether to keep, disable, or configure task_board. A configured block uses only this shape:

yaml
task_board:
  type: <registered board_type>
  project: <optional project prefix>
  key_pattern: '<optional task-key pattern>'
  create_target: <selected target id or null>
  done_target: <selected target id or null>
  options: {}
  sync_filter:
    max_age_days: <integer >= 1, or omit for no age limit>
    include_archived: <boolean, default true>

project scopes board sync and task retrieval. Explain that an empty task_board.project can mix all projects, then ask for the intended project prefix.

sync_filter is a generic sibling of provider options. The sync_filter block is optional. Never put sync_filter under options. Ask two separate questions:

  • max_age_days: choose an integer greater than or equal to 1, or no age limit.
  • include_archived: choose whether archived tasks are included; the default is true.

Age uses task last-modified time and an inclusive cutoff: a task modified exactly at the cutoff is eligible. Archive is distinct from terminal/done; include_archived: false excludes only tasks known to be archived. Age filtering runs first. Only while include_archived: false, unknown archive metadata does not itself exclude the row; an archive warning is emitted only then and only when age filtering did not already exclude the row.

Show full SKILL.md (674 more words)Show less
Editing sync_filter safely

When changing only sync_filter, use this deterministic materialization procedure:

  1. Read policy layers in precedence order: non-secret ENV/deploy task_board defaults, home:review.yml, committed .review.yml, then home:repos/<owner>/<name>.yml; stop at the selected target. Never inspect or copy credential env values. For a committed target, do not read the higher repo-home layer. For the recommended home per-repo target, include all layers.
  2. If the selected layer has a non-empty task_board mapping, use that mapping alone as the edit base. Preserve every sibling and field-attached comment already present, but do not copy or overlay omitted fields from lower layers: the selected mapping already shadows the complete lower block.
  3. If the selected layer has no task_board key, resolve only the lower layers with normal whole-block replacement, then materialize the complete lower effective non-secret task_board into the selected-layer draft. Copy type, project, key_pattern, url_template, create_target, done_target, options, every other non-secret sibling, and field-attached comments. If no lower board exists, ask for a fully configured board; never write a new partial task_board containing only the filter.
  4. If the selected layer explicitly contains null or an empty mapping, preserve that disable and do not add sync_filter. Only proceed when the user explicitly chooses to replace it with a fully configured board assembled from confirmed values; never resurrect lower fields silently.
  5. For cases 2 or 3, patch only sync_filter in the chosen or materialized block. The selected layer remains a self-contained whole-block replacement.

Because policy layers replace the whole task_board block, preserve every sibling field and comment when changing sync_filter. Repositories using the same project share one task corpus, so different retention views require different project scopes. Keep home per-repo as the recommended target for repository-specific policy. A filter change is evaluated on the next successful full sync and backfills newly eligible tasks; purge remains explicit and is never enabled by a filter change.

When a board type is selected, call the read-only discovery tool:

get_board_targets(board_type=<type>, project=<project>, provider_options=<task_board.options or {}>)

Its normalized response is {board_type, project, targets, options, warnings}. Present a pick-list of targets by label; use purposes to select create_target and done_target. For every option whose required_for contains sync, create, or finish, present its choices by label and write the selected id into task_board.options. If discovery is unavailable, empty, or returns an error, fall back to asking the user for each required generic value. Do not guess targets or options.

The resulting values are non-secret metadata. Board access is configured outside this file; do not request, display, or write credentials.

Retrieval profile

Choose one profile from tracked-file structure and write all real context_limits fields:

ProfileConditionsearch_codebase: floor / ceiling / ratio / abs_floor / candidate_pool / ann_distance_maxgraph: hops / callers_topk
tiny-utilfewer than 80 tracked Python files and one package3 / 8 / 0.60 / 0.35 / 20 / 0.651 / 20
standard80–800 files4 / 15 / 0.50 / 0.30 / 30 / 0.651 / 25
large / monorepoover 800 files or at least three large packages4 / 25 / 0.45 / 0.30 / 40 / 0.601 / 30

Map count_tasks(project) to search_tasks deterministically: < 150 → 3 / 8; 150–800 → 3 / 10; 800+ → 4 / 14. A missing tool, zero count, or unavailable corpus falls back to asking the user for small/medium/large, then uses the same mapping.

context_limits.code_section is the file budget for the task-context code section (prepare_task_context): max_files: 20, max_chunks_per_file: 1, chars_per_file: 975, max_augmented_files: 3. These four defaults are the same across all three profiles above — there is no measurement backing a per-profile split, so do not invent one. The budget unit here is the FILE, not the chunk; there is no separate character-cap key, the effective character ceiling is derived from max_files/max_chunks_per_file/chars_per_file. max_augmented_files (PRI-257) is a RESERVE of file slots inside max_files for diff paths mixed in from similar tasks (similar-diffs, the section's only augmentation source) — not a cap applied to whatever the hybrid search leaves over. Write it alongside the other real context_limits fields:

yaml
context_limits:
  search_codebase:
    floor: <profile value>
    ceiling: <profile value>
    ratio: <profile value>
    abs_floor: <profile value>
    candidate_pool: <profile value>
    ann_distance_max: <profile value>
  search_tasks:
    floor: <board-size value>
    ceiling: <board-size value>
  graph:
    hops: <profile value>
    callers_topk: <profile value>
  code_section:
    max_files: 20
    max_chunks_per_file: 1
    chars_per_file: 975
    max_augmented_files: 3

Preserve every other configuration key and ask for confirmation before writing the assembled draft.

Completion

Report old/new branches, the selected branch source, changed policy keys, selected generic targets/options, and any recommended follow-up. This skill makes configuration-only recommendations and has no index side effects.

© hashgraph-online, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in plugins/mimfort/rag_for_git/plugin/skills/configure-review of hashgraph-online/awesome-codex-plugins.

Open the folder on GitHubat commit 3e1456a

Compare with similar skills

Configure Review next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Configure Review compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Configure Review this skillhashgraph-online/awesome-codex-plugins1.3k—~3.8kAutomated safety check: PassApache-2.0
Change Trackingdotnet/efcore15k—~649Automated safety check: PassMIT
DOCX Tracked Changesmohitagw15856/pm-claude-skills1.4k—~1.4kAutomated safety check: PassMIT
Configure Channelopenclaw/openclaw392k—~946Automated safety check: PassMIT
Cost Trackingaffaan-m/ECC277k1 repos~1.3kAutomated safety check: PassMIT
Feature Trackingsickn33/agentic-awesome-skills47k1 repos~2.5kAutomated safety check: WarnMIT

Similar skills

  • Change Tracking

    dotnet/efcore

    Official

    Implementation details for EF Core change tracking. An agent skill from dotnet/efcore.

    15k GitHub stars~649 tokensUpdated yesterday
    Auto-check passed
  • DOCX Tracked Changes

    mohitagw15856/pm-claude-skills

    Produce properly-formatted tracked changes for a Word document.

    1.4k GitHub stars~1.4k tokensUpdated yesterday
    Documents & OfficeAuto-check passed
  • Configure Channel

    openclaw/openclaw

    Configure and prove a chat channel with non-interactive one-liners; secrets only as SecretRefs.

    392k GitHub stars~946 tokensUpdated today
    Auto-check passed
  • Cost Tracking

    affaan-m/ECC

    Track and report Claude Code token usage, spending, and budgets from the local ECC cost-tracker metrics log.

    277k GitHub starsUsed in 1 repo~1.3k tokens
    AI & LLM EngineeringAuto-check passed
  • Feature Tracking

    sickn33/agentic-awesome-skills

    Maintain durable feature-level memory across AI coding sessions with lightweight Markdown tracks for status, source-of-truth docs, decisions, risks, and changes.

    47k GitHub starsUsed in 1 repo~2.5k tokens
    Auto-check: warnings
  • Branches

    redis/RedisInsight

    Official

    Create and name git branches following project conventions. An agent skill from redis/RedisInsight.

    8.9k GitHub stars~409 tokensUpdated 6 days ago
    DevelopmentAuto-check passed

More from hashgraph-online/awesome-codex-plugins

All 715 skills in this repo
  • Anime Reaction Gif

    hashgraph-online/awesome-codex-plugins

    Create original anime-style reaction stickers as looping GIFs and MP4 previews, using generated character pose sheets and timed key poses.

    1.3k GitHub stars~922 tokensUpdated today
    Auto-check passed
  • Calibredb

    hashgraph-online/awesome-codex-plugins

    Manage and query Calibre libraries with the calibredb CLI (local paths or Calibre Content server URLs).

    1.3k GitHub stars~1k tokensUpdated today
    Auto-check passed
  • Rust API Test Harness

    hashgraph-online/awesome-codex-plugins

    A skill your agent uses when adding, changing, testing, or debugging Rust HTTP APIs and services, especially when Codex needs black-box integration tests, random-port app startup, real database test…

    1.3k GitHub stars~1.7k tokensUpdated today
    Auto-check passed
  • Art

    hashgraph-online/awesome-codex-plugins

    Make a studio's game look like something at build time — a cover from a real frame of the game (free), painted covers, backdrops, textures and character plates from image models through the…

    1.3k GitHub stars~2.5k tokensUpdated today
    Auto-check passed
  • Calle

    hashgraph-online/awesome-codex-plugins

    Use CALL-E from Codex through the calle CLI. An agent skill from hashgraph-online/awesome-codex-plugins.

    1.3k GitHub stars~2.9k tokensUpdated today
    Auto-check passed
  • Game Balance Economy

    hashgraph-online/awesome-codex-plugins

    Balance game difficulty, resources, rewards, probability, progression, economies, and dominant strategies.

    1.3k GitHub stars~618 tokensUpdated today
    Auto-check passed

Questions about Configure Review

What does Configure Review do?

A skill your agent uses when configuring or changing a repository's tracked branches, layered review policy, ignored tracked paths, retrieval limits, summary depth, or non-secret task-board metadata. Configure Review is an agent skill from hashgraph-online/awesome-codex-plugins. Use when configuring or changing a repository's tracked branches, layered review policy, ignored tracked paths, retrieval limits, summary depth, or non-secret task-board metadata.

When should I use Configure Review?

Configure Review fits situations like: changing a repositorys tracked branches; layered review policy; ignored tracked paths; retrieval limits.

How do I install Configure Review in Claude Code?

Run `npx skills add hashgraph-online/awesome-codex-plugins --skill configure-review -a claude-code`. Or copy the skill folder (plugins/mimfort/rag_for_git/plugin/skills/configure-review in hashgraph-online/awesome-codex-plugins) into .claude/skills/configure-review in your project. Claude Code loads it when a task matches its description.

How do I install Configure Review in Codex?

Run `npx skills add hashgraph-online/awesome-codex-plugins --skill configure-review -a codex`. Or copy the skill folder (plugins/mimfort/rag_for_git/plugin/skills/configure-review in hashgraph-online/awesome-codex-plugins) into .agents/skills/configure-review in your project. Codex loads it when a task matches its description.

Can I use Configure Review in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add hashgraph-online/awesome-codex-plugins --skill configure-review -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/configure-review, .gemini/skills/configure-review, .github/skills/configure-review and .opencode/skills/configure-review in your project.

What does Configure Review need to run?

Going by SKILL.md and its folder, Configure Review needs the command-line tools its instructions call (git). Our summary lists: Python 3.

Does Configure Review access the network?

SKILL.md contains no URLs. Its commands use git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Configure Review safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Configure Review use?

Configure Review is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Configure Review use?

About 3.8k tokens (SKILL.md is roughly 15k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Configure Review?

Skills that share tags, products or a category with Configure Review: Change Tracking (dotnet/efcore, 15k stars), DOCX Tracked Changes (mohitagw15856/pm-claude-skills, 1.4k stars), Configure Channel (openclaw/openclaw, 392k stars) and Cost Tracking (affaan-m/ECC, 277k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Configure Review?

hashgraph-online (a GitHub organization) maintains it in hashgraph-online/awesome-codex-plugins, which has 1,267 GitHub stars. The repository holds 716 skills in this directory. The repository was last updated on October 10, 2026.

Source: hashgraph-online/awesome-codex-plugins on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.