New: 98,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Review

by akseolabs-seo in akseolabs-seo/AK-Threads-booster

Post-publish feedback loop: collect actual metrics, compare against predictions, update the tracker, refresh style conclusions carefully, and learn from deviations.

MITAuto-check passedWriting & Content

Install Review

skills CLI
$ npx skills add akseolabs-seo/AK-Threads-booster --skill review -a claude-code

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

GitHub CLI
$ gh skill install akseolabs-seo/AK-Threads-booster 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/akseolabs-seo/AK-Threads-booster.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/review .claude/skills/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
review
GitHub stars
275
Token cost
~1.9k tokens
SKILL.md length
963 words
Files
5 (incl. references)
Skills in repo
12
Repo updated
First seen
Licence
MIT

At a glance

Post-publish feedback loop: collect actual metrics, compare against predictions, update the tracker, refresh style conclusions carefully, and learn from deviations.

  • Works in 12 steps: Sweep Expired Prediction Placeholders → Collect Actual Data → Compare Prediction vs Actual → …
  • Writing & Content work in your project
  • SKILL.md covers Principles & Knowledge, User Data Paths, Execution Flow and Boundary Reminders
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Review is an agent skill from akseolabs-seo/AK-Threads-booster. Post-publish feedback loop: collect actual metrics, compare against predictions, update the tracker, refresh style conclusions carefully, and learn from deviations.

Its SKILL.md is about 1.9k tokens, which your agent loads only when the skill is triggered. The skill folder holds 5 other files, including reference files (for example `references/log-hygiene.md`, `references/output-format.md` and `references/skill-learning-capture.md`).

It sits in Writing & Content. The repository describes itself as: AK體 · 數據驅動的 Threads 寫文決策系統。用你的歷史貼文、演算法與社媒心理學，協助選題、起草、發文前診斷、表現預估與復盤。Data-driven Threads writing advisor — topic selection, drafting, diagnosis, prediction & review based on your… The licence is MIT.

When your agent uses it

  • Writing & Content work in your project

Example prompts

  • “/review”

Requirements

  • Pre-approved tools (allowed-tools): Read, Write, Edit, Grep, Glob

Workflow steps

12 steps, taken from the step headings in SKILL.md.

  1. Sweep Expired Prediction Placeholders
  2. Collect Actual Data
  3. Compare Prediction vs Actual
  4. Deviation Analysis
  5. 5: Backup Before Write
  6. Update Tracker
  7. Refresh Style Guide Carefully
  8. Update Concept Library
  9. 4: Rebuild Compiled Memory
  10. 5 + 6.6: Log-Hygiene Checks
  11. Output
  12. Skill-Level Learning Capture (opt-in, non-blocking)

What it can do on your machine

Read from SKILL.md and the folder at commit cc08954. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves these tools, so the agent can use them without asking each time:

    • Read
    • Write
    • Edit
    • Grep
    • Glob

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

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

  • Network

    No URLs in SKILL.md.

    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

Review loads about 1.9k tokens when it runs, and up to ~4.7k if it reads all its reference files. Until then it costs about 43 tokens; SKILL.md has 963 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~43
When it runs · the whole SKILL.md, loaded when a task matches
~1.9k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~4.7k

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 akseolabs-seo/AK-Threads-booster at commit cc08954, republished under its MIT licence (© akseolabs-seo). 963 words, ~1,943 tokens.

Download SKILL.mdSave it as .claude/skills/review/SKILL.md (or your agent's skills folder). This skill also uses 4 other files; get the full folder from GitHub.
name
review
description
Post-publish feedback loop: collect actual metrics, compare against predictions, update the tracker, refresh style conclusions carefully, and learn from deviations.
allowed-tools
Read, Write, Edit, Grep, Glob
version
2.0.0

AK-Threads-Booster Post-Publish Feedback Module (M8 + M9)

You are the data feedback consultant for the AK-Threads-Booster system. After a post is published, collect actual performance data, compare it with prior expectations, and update the data assets cautiously.

Principles & Knowledge

Load knowledge/_shared/principles.md before running feedback. Follow discovery order in knowledge/_shared/discovery.md. For /review specifically, load _shared/config.md, _shared/runtime-budget.md, _shared/next-move-engine.md, algorithm-card.md, and data-confidence.md.

Load full algorithm.md only in deep mode or when the outcome deviation depends on an ambiguous algorithm interpretation.

Skill-specific addendum: prediction error is normal — the job is to learn why, not to score the user. One post should not override a stable historical trend.

User Data Paths

Search for:

  • threads_daily_tracker.json
  • compiled/account_wiki.md
  • compiled/account_state.md
  • compiled/personal_signal_memory.md
  • compiled/next_move_queue.md
  • compiled/post_feature_index.jsonl
  • compiled/cluster_wiki.json
  • compiled/recent_window.md
  • style_guide.md
  • concept_library.md

If the tracker is missing, tell the user to supply historical data or run /setup first.

Before loading broader history or algorithm context, resolve runtime.token_mode per knowledge/_shared/runtime-budget.md. If absent or "ask", ask whether this run should use low-token or high-token mode and show the pros/cons. Low-token is enough for routine prediction-vs-actual review; high-token is better when the deviation is surprising or strategically important.

Execution Flow

Step 0: Sweep Expired Prediction Placeholders

Walk posts[] and find entries where id starts with pending- and pending_expires_at is earlier than now.

For each match:

  1. If the user is present, ask once whether to discard (draft was never published) or extend (still planning).
  2. On discard, move the entry to discarded_drafts[] at the tracker root (create if missing) with a discarded_at timestamp and the original prediction_snapshot. Do not delete outright — the prediction itself is a learning signal.
  3. On extend, push pending_expires_at forward by 7 days.

In headless contexts (no user), default to discard. This keeps /topics, /analyze, and data-confidence counts from being polluted by abandoned drafts.

Step 1: Collect Actual Data

Method A — User-provided metrics. The user supplies: which post, hours after publish, views, likes, replies, reposts, shares.

Method B — Tracker-backed metrics. Read existing tracker data and update the relevant performance window if newer data is available. If the user has API access, prefer a tracker kept fresh via scripts/update_snapshots.py — it appends snapshots[] and updates the closest performance_windows checkpoint automatically.

Step 2: Compare Prediction vs Actual

If posts[i].prediction_snapshot exists, build the comparison table and play-out notes per references/output-format.md (Prediction-vs-actual section).

If no prediction_snapshot exists, skip this section cleanly and say so. Do not invent a prior prediction.

Step 3: Deviation Analysis

If the review identifies a next-post direction, use knowledge/_shared/next-move-engine.md: recommend the next move in plain Chinese, name the S signal it should strengthen, and name the R risks to avoid. Do not turn the review into a formula prescription.

Walk the deviation-analysis checklist in references/tracker-update-fields.md. Phrase findings as observations, not verdicts ("may relate to…, for your reference").

Step 3.5: Backup Before Write

Follow the destructive-writes policy in templates/FAILSAFE.md. Before mutating any of threads_daily_tracker.json, style_guide.md, or concept_library.md:

  1. Back up each file to <filename>.bak-<ISO> (compact UTC ISO, e.g. 20260418T143012Z).
  2. If any backup fails, abort the entire review-update phase and tell the user which file failed. No partial writes across these three files.
  3. Write to a .tmp-<ISO> sibling, then atomically rename over the target.
  4. Prune older backups, keeping at most 5 per file.

Reason: /review is the most destructive sub-skill. The FAILSAFE policy is centralized so every write-capable sub-skill (/predict, /refresh, /voice, /setup) honors the same contract.

Show full SKILL.md (410 more words)Show less
Step 4: Update Tracker

Update only the fields listed in references/tracker-update-fields.md (post-level, algorithm signals, psychology signals, snapshot/windows, review state, top-level). Do not break the schema. Preserve existing fields.

prediction_snapshot is owned exclusively by /predict — do not write or overwrite it from /review. If a prediction needs to be recorded after the fact, ask the user to re-run /predict.

Step 5: Refresh Style Guide Carefully

Update style_guide.md only when the new post adds a meaningful data point on one of the dimensions listed in references/tracker-update-fields.md (style-guide refresh scope). One post can extend a trend; it should not overturn a stable trend by itself.

Step 6: Update Concept Library

If the post introduced new concepts or analogies, add them to concept_library.md with explanation depth and a note on whether the analogy is reusable or overused.

Step 6.4: Rebuild Compiled Memory

After tracker/style/concept updates succeed, rebuild compiled memory with scripts/build_compiled_memory.py --tracker ./threads_daily_tracker.json. If this fails, keep the completed review updates and report that low-token runtime is stale until the script is rerun.

Step 6.5 + 6.6: Log-Hygiene Checks

Run the freshness-log and refresh-log hygiene checks per references/log-hygiene.md. Both are advisory — surface findings in the Step 7 report but never block the review.

Step 7: Output

Produce the Post-Publish Feedback Report exactly per references/output-format.md. Omit subsections cleanly when the underlying data does not exist — never invent placeholders.

Step 8: Skill-Level Learning Capture (opt-in, non-blocking)

See references/skill-learning-capture.md for the full trigger condition, append procedure, and ≥10-entry threshold message. Key rules in one paragraph:

Only write to threads_skill_learnings.log when the user explicitly confirms a skill-level miss in this session — a verbatim user_signal quote is required. Follow the schema in knowledge/_shared/compound-log-format.md and the append-only policy in templates/FAILSAFE.md. Never auto-patch sub-skills. When the log reaches ≥10 entries, surface a one-line pointer to /optimize in the Cumulative Learning section — /optimize ships with this skill and requires user approval per proposed edit.

If the user declines the capture or doesn't signal a miss, skip this step silently.

Boundary Reminders

  • If no prior prediction exists, skip prediction comparison cleanly.
  • If the tracker is partial-data only, say which conclusions remain weak.
  • If there is no API-backed snapshot flow, use checkpoint data only. Do not pretend to have a growth curve.
  • Keep updates cumulative and reversible in logic.
  • When discovery-surface data is unavailable, say so explicitly instead of inferring a source mix with false precision.
  • Compiled memory is a cache. Never let it override tracker data during review; rebuild it after successful tracker mutations.

© akseolabs-seo, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 4 other files (references) in skills/review of akseolabs-seo/AK-Threads-booster.

  • SKILL.md
  • references/log-hygiene.md
  • references/output-format.md
  • references/skill-learning-capture.md
  • references/tracker-update-fields.md

Open the folder on GitHubat commit cc08954

Compare with similar skills

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.

Review compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Review this skillakseolabs-seo/AK-Threads-booster275—~1.9kAutomated safety check: PassMIT
Socialcoreyhaines31/marketingskills54k4 repos~4.5kAutomated safety check: PassMIT
HumanizerAzure-Samples/interview-coach-agent-framework17338 repos~5.8kAutomated safety check: PassMIT
Avoid AI Writingconorbronsdon/avoid-ai-writing4.9k3 repos~8.1kAutomated safety check: PassMIT
JavaScript Concept Fact Checkerleonardomso/33-js-concepts67k1 repos~5kAutomated safety check: PassMIT
User-Facing Text Cleanupguillaumemeyer/watermarks-remover24k—~3.5kAutomated safety check: PassMIT

Similar skills

  • Social

    coreyhaines31/marketingskills

    When the user wants help creating, scheduling, or optimizing social media content for LinkedIn, Twitter/X, Instagram, TikTok, or Facebook, or wants to do social listening and engagement triage.

    54k GitHub starsUsed in 4 repos~4.5k tokens
    Writing & ContentAuto-check passed

  • Humanizer

    Azure-Samples/interview-coach-agent-framework

    Official

    Remove signs of AI-generated writing from text. An agent skill from Azure-Samples/interview-coach-agent-framework.

    173 GitHub starsUsed in 38 repos~5.8k tokens
    Writing & ContentAuto-check passed

  • Avoid AI Writing

    conorbronsdon/avoid-ai-writing

    Audit and rewrite content to remove AI writing patterns ("AI-isms").

    4.9k GitHub starsUsed in 3 repos~8.1k tokens
    Writing & ContentAuto-check passed

  • JavaScript Concept Fact Checker

    leonardomso/33-js-concepts

    Verifies the technical accuracy of JavaScript concept pages by checking code examples, MDN and ECMAScript claims and external links through a five-phase method.

    67k GitHub starsUsed in 1 repo~5k tokens
    Writing & ContentAuto-check passed

  • User-Facing Text Cleanup

    guillaumemeyer/watermarks-remover

    Audits prose for invisible Unicode characters and rewrites it while keeping facts, citations, code and required disclosures unchanged and the writer's voice intact.

    24k GitHub stars~3.5k tokensUpdated 2 days ago
    Writing & ContentAuto-check passed

  • Install Anti Slop

    trycompai/crm

    Install and configure the anti-slop Oxlint plugin in a local TypeScript or JavaScript repository.

    11k GitHub starsUsed in 1 repo~881 tokens
    Writing & ContentAuto-check passed

More from akseolabs-seo/AK-Threads-booster

All 12 skills in this repo

  • Voice

    akseolabs-seo/AK-Threads-booster

    Deep analysis of user's historical posts and comment replies to build a comprehensive Brand Voice profile.

    275 GitHub stars~2k tokensUpdated 3 mo ago
    Auto-check: notes

  • Optimize

    akseolabs-seo/AK-Threads-booster

    Self-contained compound loop: read threadsskilllearnings.log, cluster the misses, propose concrete sub-skill rule edits, and apply them with the user's approval.

    275 GitHub stars~1.8k tokensUpdated 3 mo ago
    Auto-check passed

  • Panel

    akseolabs-seo/AK-Threads-booster

    Launch or prepare the optional local visual panel for AK-Threads-Booster.

    275 GitHub stars~873 tokensUpdated 3 mo ago
    Auto-check: notes

  • Topics

    akseolabs-seo/AK-Threads-booster

    Mine insights from comments and historical data to recommend the next worthwhile topics.

    275 GitHub stars~1.8k tokensUpdated 3 mo ago
    Auto-check: notes

  • Update

    akseolabs-seo/AK-Threads-booster

    Check AK-Threads-Booster for upstream GitHub updates, safely fast-forward the local skill repo, or install an opt-in weekly Codex automation that keeps the skill on the latest version.

    275 GitHub stars~1.1k tokensUpdated 3 mo ago
    Auto-check: notes

  • Draft

    akseolabs-seo/AK-Threads-booster

    Select a topic and generate a draft based on the user's Brand Voice.

    275 GitHub stars~2.6k tokensUpdated 3 mo ago
    Auto-check passed

Categories

Questions about Review

What does Review do?

Post-publish feedback loop: collect actual metrics, compare against predictions, update the tracker, refresh style conclusions carefully, and learn from deviations. Review is an agent skill from akseolabs-seo/AK-Threads-booster. Post-publish feedback loop: collect actual metrics, compare against predictions, update the tracker, refresh style conclusions carefully, and learn from deviations.

When should I use Review?

Review fits situations like: writing & Content work in your project.

How do I install Review in Claude Code?

Run `npx skills add akseolabs-seo/AK-Threads-booster --skill review -a claude-code`. Or copy the skill folder (skills/review in akseolabs-seo/AK-Threads-booster) into .claude/skills/review in your project. Claude Code loads it when a task matches its description.

How do I install Review in Codex?

Run `npx skills add akseolabs-seo/AK-Threads-booster --skill review -a codex`. Or copy the skill folder (skills/review in akseolabs-seo/AK-Threads-booster) into .agents/skills/review in your project. Codex loads it when a task matches its description.

Can I use 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 akseolabs-seo/AK-Threads-booster --skill 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/review, .gemini/skills/review, .github/skills/review and .opencode/skills/review in your project.

What does Review need to run?

SKILL.md names no scripts, command-line tools or credentials: Review is instructions for the agent only. Its frontmatter pre-approves these tools: Read, Write, Edit, Grep, Glob.

Does Review access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is 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 Review use?

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

How many tokens does Review use?

About 1.9k tokens (SKILL.md is roughly 7.8k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 2.7k tokens, read only when the agent opens those files.

What are the alternatives to Review?

Skills that share tags, products or a category with Review: Social (coreyhaines31/marketingskills, 54k stars), Humanizer (Azure-Samples/interview-coach-agent-framework, 173 stars), Avoid AI Writing (conorbronsdon/avoid-ai-writing, 4.9k stars) and JavaScript Concept Fact Checker (leonardomso/33-js-concepts, 67k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Review?

akseolabs-seo (a GitHub user) maintains it in akseolabs-seo/AK-Threads-booster, which has 275 GitHub stars. The repository holds 12 skills in this directory. The repository was last updated on July 3, 2026.

Source: akseolabs-seo/AK-Threads-booster on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.