New: 97,000+ 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

Ablate AI Layer

by coleam00 in coleam00/skills

Measure whether a repository's AI instructions still earn their place, by running the same real task many times with the layer intact and with it stripped, then grading every rule against what…

MITAuto-check passedAgent Workflows

Install Ablate AI Layer

skills CLI
$ npx skills add coleam00/skills --skill ablate-ai-layer -a claude-code

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

GitHub CLI
$ gh skill install coleam00/skills ablate-ai-layer --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/coleam00/skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/ablate-ai-layer .claude/skills/ablate-ai-layer && 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
ablate-ai-layer
GitHub stars
670
Token cost
~1.9k tokens
SKILL.md length
1,080 words
Files
4 (incl. scripts, references)
Skills in repo
33
Repo updated
First seen
Licence
MIT

At a glance

Measure whether a repository's AI instructions still earn their place, by running the same real task many times with the layer intact and with it stripped, then grading every rule against what…

  • Works in 7 steps: Map the layer → Get the probe task → Show the plan and get approval → …
  • The user wants to prune
  • SKILL.md covers What makes the result…, Step 1. Map the layer, Step 2. Get the probe task and Step 3. Show the plan and get…, plus 6 more sections
  • Runs Python scripts from its folder; calls python

What it does

Ablate AI Layer is an agent skill from coleam00/skills. Measure whether a repository's AI instructions still earn their place, by running the same real task many times with the layer intact and with it stripped, then grading every rule against what actually changed. Runs both arms itself in throwaway git worktrees and never touches the working tree. Agent-agnostic across CLAUDE.md, AGENTS.md, .claude/, .agents/, .cursor/rules, .clinerules, .windsurfrules and copilot-instructions. Use when the user wants to prune, audit, clean up, shrink or "delete" their CLAUDE.md…

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 scripts and reference files (for example `references/comparison.md`, `scripts/map_layer.py` and `scripts/run_ablation.py`).

It sits in Agent Workflows, covering Agent instruction files. The repository describes itself as: The agent skills I actually use to build software with coding agents. The PIV loop, planning, worktrees, and the meta-skills for building your own AI Layer. The licence is MIT.

When your agent uses it

  • The user wants to prune
  • Delete their CLAUDE.md
  • Agent instructions
  • They ask whether their rules are still needed

Example prompts

  • “delete”
  • “/ablate-ai-layer”

Requirements

  • Python 3

Workflow steps

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

  1. Map the layer
  2. Get the probe task
  3. Show the plan and get approval
  4. Run it
  5. Grade
  6. Report
  7. Apply, with the user's approval

What it can do on your machine

Read from SKILL.md and the folder at commit dfaa910. 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

    Ships 2 files in scripts/ (Python), which the agent can run.

    Shell commands in SKILL.md call:

    • python

    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

Ablate AI Layer loads about 1.9k tokens when it runs, and up to ~3.4k if it reads all its reference files. Until then it costs about 195 tokens; SKILL.md has 1,080 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~195
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
~3.4k

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); the scripts in this folder are not scanned.

SKILL.md

The full file from coleam00/skills at commit dfaa910, republished under its MIT licence (© coleam00). 1,080 words, ~1,914 tokens.

Download SKILL.mdSave it as .claude/skills/ablate-ai-layer/SKILL.md (or your agent's skills folder). This skill also uses 3 other files; get the full folder from GitHub.
name
ablate-ai-layer
description
Measure whether a repository's AI instructions still earn their place, by running the same real task many times with the layer intact and with it stripped, then grading every rule against what actually changed. Runs both arms itself in throwaway git worktrees and never touches the working tree. Agent-agnostic across CLAUDE.md, AGENTS.md, .claude/, .agents/, .cursor/rules, .clinerules, .windsurfrules and copilot-instructions. Use when the user wants to prune, audit, clean up, shrink or "delete" their CLAUDE.md, AGENTS.md, cursor rules, agent instructions or AI layer; when they ask whether their rules are still needed, whether their context is bloated, or what to cut; or when they mention ablating, ablation, or testing their agent without its instructions.

Ablate the AI layer

Model upgrades quietly retire instructions. A rule written to work around a weaker model becomes dead weight that competes for attention with the rules that still matter. Reading the file will not tell you which is which. Only an experiment will.

You run the experiment. The user picks the task and approves the conclusion. Do not hand the user a list of commands to run; the script drives both arms.

What makes the result trustworthy

  • Both arms, many runs each. A stripped agent does not visibly fail, so a single run has nothing to compare against and "seems fine" becomes "delete something load-bearing". Two runs of the same arm can also differ more than the two arms differ, so one pair per arm is the floor, not the target.
  • Nothing is moved aside. Every run happens in a detached git worktree built from HEAD in a temp directory, outside the repo, and deleted afterwards. The user's working tree is never modified, so there is no restore step to forget.
  • Only the always-loaded set is stripped by default. Skills, subagents and path-scoped rules cost nothing until they fire, so deleting them buys back no context. Hooks and permissions are never touched: they run as code and spend no attention.

Step 1. Map the layer

bash
python <skill>/scripts/map_layer.py [repo_root]

Read-only. Sorts every artifact into always-loaded, on-demand, and enforcement, and prints what the always-loaded set costs on every session before the user types anything. Show them that number.

If nothing is found, say so and stop. There is nothing to test.

Step 2. Get the probe task

This is the one thing you must not decide for the user. The task determines whether the experiment can detect anything at all.

A good probe task is real work they would do anyway, touches code where house conventions plausibly apply, and adds something that has to be wired in: a test, an endpoint, a migration, a command.

A bad one is a typo, a rename, or any one-line fix. It is fully derivable, both arms will match, and the user will wrongly conclude their whole layer is worthless. Say that out loud if they offer one, and ask for something with conventions at stake.

Write the agreed task verbatim to a file. Every run reuses it byte for byte.

Step 3. Show the plan and get approval

Report before spending anything: how many runs, which model, roughly what it will cost, and that the working tree will not be touched.

bash
python <skill>/scripts/run_ablation.py <repo> --task-file <task.md> --dry-run

The dry run also surfaces two things worth pausing on:

  • A dirty working tree. Worktrees are built from HEAD, so uncommitted edits are not under test. Offer to commit or stash first.
  • A build-dependency warning. Some repos import their own AI layer as source. A CLI that reads its skill markdown at build time breaks the moment those files go missing, and the user will read a compile error as an agent regression. Keep --scope always if this warns.

Step 4. Run it

bash
python <skill>/scripts/run_ablation.py <repo> --task-file <task.md> --runs 2

This is the whole experiment. It builds a fresh worktree per run, strips the layer in the stripped arm, runs the same prompt in each, captures every diff, cleans up every worktree, and writes results to .ablation/<timestamp>/ (gitignored).

Useful flags: --runs 3 when the user intends to act on the result, --scope all to test the harder claim that skills and subagents have expired too, --model, --jobs for concurrency, --runner for a non-Claude agent that reads a prompt on stdin.

If an arm produced nothing usable, stop. An empty arm is a broken experiment, not a finding. Re-run before drawing anything from it.

Show full SKILL.md (479 more words)Show less

Step 5. Grade

Read references/comparison.md before analysing. It is the rubric, and it contains the two things that make the difference between a real result and a confident wrong one: grade per rule rather than diffing the arms against each other, and grade blind to which arm a diff came from.

The short version:

  1. Turn the always-loaded files into a numbered checklist of testable claims. Mark anything unfalsifiable ("write clean code") as exactly that.
  2. Judge every run's diff against every claim: followed, violated, or n/a.
  3. Only then join verdicts back to arms and read the pattern.

Step 6. Report

Give the user a table, one row per rule, sorted so the actionable rows are first:

Pattern across runsVerdictAction
control follows, stripped violatesload-bearingkeep, rewrite shorter
both arms followmodel does this anywaydelete
both arms violateignored even when loadedmake it a hook or test, or delete
never applicableuntestedkeep, no evidence either way
inconsistent within an armnoisemore runs or a better task

Keep "untested" visually separate from "no difference". They look identical in the data and mean opposite things, and merging them is how a rule that protects a case this task never touched gets deleted.

Step 7. Apply, with the user's approval

Never edit the rules file unattended. Propose the edit, show the diff, wait.

Re-add or keep one line at a time, only for rules with observed evidence, and prefer a test, then a hook, then an on-demand instruction, and only then an always-loaded line. Finish by re-running map_layer.py so the new always-loaded total sits next to the old one.

Honest framing to give the user

  • One probe task is a data point, not a verdict. Encourage a second task on a different part of the codebase before deleting anything large.
  • A null result is a real result. If the arms match, that part of the layer has genuinely expired and can go.
  • The reverse is also true. Do not let one clean run justify deleting rules for cases this task never exercised: security, compliance, release procedure.
  • Existing code substitutes for the rules file. A stripped run copies conventions from neighbouring code when there is a neighbour to copy. The same rule can hold in an edited file and break in a new one. Weight new-file evidence more heavily and say which kind each verdict rests on.
  • Cheaper and smaller models lean on instructions more than frontier models do. A layer that looks redundant under a frontier model may still be carrying a cheaper one. If the team runs a mix, ablate against the weakest model in use.

Resources

  • scripts/map_layer.py: read-only inventory of the layer, agent-agnostic.
  • scripts/run_ablation.py: runs both arms and collects the diffs. --help lists every flag. Never read either script into context; only their output.
  • references/comparison.md: the grading rubric. Read it before Step 5.

© coleam00, 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 3 other files (scripts, references) in .claude/skills/ablate-ai-layer of coleam00/skills.

  • SKILL.md
  • references/comparison.md
  • scripts/map_layer.py
  • scripts/run_ablation.py

Open the folder on GitHubat commit dfaa910

Compare with similar skills

Ablate AI Layer 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.

Ablate AI Layer compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Ablate AI Layer this skillcoleam00/skills670—~1.9kAutomated safety check: PassMIT
Using Agent Skillsaddyosmani/agent-skills102k4 repos~2.4kAutomated safety check: PassMIT
Claude ReflectBayramAnnakov/claude-reflect1.7k2 repos~627Automated safety check: PassMIT
Neat-Freak Knowledge CloseoutKKKKhazix/khazix-skills21k—~1.9kAutomated safety check: PassMIT
Task Observerrebelytics/one-skill-to-rule-them-all3.2k1 repos~12kAutomated safety check: PassCC-BY-4.0
Writing For Agentsbestofjs/bestofjs3.1k17 repos~2.7kAutomated safety check: PassMIT

Similar skills

  • Using Agent Skills

    addyosmani/agent-skills

    Meta-skill for choosing which workflow skill fits the task at hand, plus always-on habits: surface assumptions, stop on confusion, push back, keep it simple and stay in scope.

    102k GitHub starsUsed in 4 repos~2.4k tokens
    Agent WorkflowsAuto-check passed

  • Claude Reflect

    BayramAnnakov/claude-reflect

    Self-learning system that captures corrections during sessions and reminds users to run /reflect to update CLAUDE.md.

    1.7k GitHub starsUsed in 2 repos~627 tokens
    Agent WorkflowsAuto-check passed

  • Neat-Freak Knowledge Closeout

    KKKKhazix/khazix-skills

    Brings project docs, agent rule files, authorized memory and leftover workspace files back in line with what the code and runtime actually do at the end of a work session.

    21k GitHub stars~1.9k tokensUpdated 6 days ago
    Agent WorkflowsAuto-check passed

  • Task Observer

    rebelytics/one-skill-to-rule-them-all

    Monitors task execution for skill improvement opportunities.

    3.2k GitHub starsUsed in 1 repo~12k tokens
    Agent WorkflowsAuto-check passed

  • Writing For Agents

    bestofjs/bestofjs

    Writing documents for agents. An agent skill from bestofjs/bestofjs.

    3.1k GitHub starsUsed in 17 repos~2.7k tokens
    Agent WorkflowsAuto-check passed

  • SkillOpt Sleep Cycle

    microsoft/SkillOpt

    Official

    Runs an on-demand or nightly sleep cycle that reviews past Claude Code sessions and proposes validated updates to CLAUDE.md and skills.

    18k GitHub stars~2.3k tokensUpdated yesterday
    Agent WorkflowsAuto-check passed

More from coleam00/skills

All 33 skills in this repo

  • Build Dark Factory

    coleam00/skills

    Take a PRD and build a dark factory around it - a repository that takes work in as an issue and ships validated code out with nobody at the keyboard - one component at a time, into the user's actual…

    670 GitHub stars~12k tokensUpdated 20 days ago
    Auto-check passed

  • Drive Screen

    coleam00/skills

    Take real control of the desktop - list and focus windows, type, paste, click, scroll, and screenshot - on Windows, macOS or Linux, and drive other coding-agent sessions running in terminals.

    670 GitHub stars~5.4k tokensUpdated 20 days ago
    Auto-check passed

  • Second Brain Audit

    coleam00/skills

    Audit any second brain, notes folder, or agent memory for facts that have quietly stopped being true, then fix the worst one so it stops recurring.

    670 GitHub stars~4.6k tokensUpdated 20 days ago
    Auto-check passed

  • Worktree Create

    coleam00/skills

    Create one or more git worktrees for parallel development, each on its own branch with gitignored config copied in, dependencies installed, and a health check, by fanning out a setup subagent per…

    670 GitHub stars~958 tokensUpdated 20 days ago
    Auto-check passed

  • Hooks Create

    coleam00/skills

    Author a working Claude Code hook from a plain-English description of what it should guarantee or do.

    670 GitHub stars~3.6k tokensUpdated 20 days ago
    Auto-check: notes

  • Piv Implement

    coleam00/skills

    Executes an implementation plan task-by-task with validation at every step.

    670 GitHub stars~1k tokensUpdated 20 days ago
    Auto-check passed

Categories

Questions about Ablate AI Layer

What does Ablate AI Layer do?

Measure whether a repository's AI instructions still earn their place, by running the same real task many times with the layer intact and with it stripped, then grading every rule against what…. Ablate AI Layer is an agent skill from coleam00/skills. Measure whether a repository's AI instructions still earn their place, by running the same real task many times with the layer intact and with it stripped, then grading every rule against what actually changed.

When should I use Ablate AI Layer?

Ablate AI Layer fits situations like: the user wants to prune; delete their CLAUDE.md; agent instructions; they ask whether their rules are still needed.

How do I install Ablate AI Layer in Claude Code?

Run `npx skills add coleam00/skills --skill ablate-ai-layer -a claude-code`. Or copy the skill folder (.claude/skills/ablate-ai-layer in coleam00/skills) into .claude/skills/ablate-ai-layer in your project. Claude Code loads it when a task matches its description.

How do I install Ablate AI Layer in Codex?

Run `npx skills add coleam00/skills --skill ablate-ai-layer -a codex`. Or copy the skill folder (.claude/skills/ablate-ai-layer in coleam00/skills) into .agents/skills/ablate-ai-layer in your project. Codex loads it when a task matches its description.

Can I use Ablate AI Layer 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 coleam00/skills --skill ablate-ai-layer -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/ablate-ai-layer, .gemini/skills/ablate-ai-layer, .github/skills/ablate-ai-layer and .opencode/skills/ablate-ai-layer in your project.

What does Ablate AI Layer need to run?

Going by SKILL.md and its folder, Ablate AI Layer needs Python for the scripts in its folder and the command-line tools its instructions call (python). Our summary lists: Python 3.

Does Ablate AI Layer 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 Ablate AI Layer 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. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Ablate AI Layer use?

Ablate AI Layer 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 Ablate AI Layer use?

About 1.9k tokens (SKILL.md is roughly 7.7k 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 1.5k tokens, read only when the agent opens those files.

What are the alternatives to Ablate AI Layer?

Skills that share tags, products or a category with Ablate AI Layer: Using Agent Skills (addyosmani/agent-skills, 102k stars), Claude Reflect (BayramAnnakov/claude-reflect, 1.7k stars), Neat-Freak Knowledge Closeout (KKKKhazix/khazix-skills, 21k stars) and Task Observer (rebelytics/one-skill-to-rule-them-all, 3.2k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Ablate AI Layer?

coleam00 (a GitHub user) maintains it in coleam00/skills, which has 670 GitHub stars. The repository holds 33 skills in this directory. The repository was last updated on September 17, 2026.

Source: coleam00/skills on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.