skillet
llms.txt13 starsChanged 7 months agoMITView on GitHub
Build agent skills from a reviewable spec, with optional evals for selected behavior. Skillet keeps the required spec and agent instructions reviewable. Eval cases are a separate opt-in for scenarios worth measuring; ordinary authoring does not create or run them. When asked to install Skillet globally for an agent: The latest skillet status checks the standard user-scoped installation for a stale authoring contract. When stale, follow status.next: reinstall through its scoped dotagents command or the same method that originally installed…
In this repository16 other files it gives its agents.
- AGENTS.mdAGENTS.md
- examples/garfield/evals/fixtures/review-slice/AGENTS.mdAGENTS.md
- openspec-apply-changeSkill
- openspec-archive-changeSkill
- openspec-exploreSkill
- openspec-proposeSkill
- commit-conventionsSkill
- effectSkill
- effectSkill
- garfieldSkill
- garfieldSkill
- safe-deleteSkill
- safe-deleteSkill
- todo-hygieneSkill
- safe-deleteSkill
- skillet-authoringSkill
The file
We split this llms.txt into the rules it gives an agent
4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 13
When asked to install Skillet globally for an agent:
- Distinctive: line 17
Do not create project-local skill files or modify the current repository.
- Very specific: line 18
After installation, run Skillet commands through npx -y @sentry/skillet@latest or pnpx @sentry/skillet@latest.
- Distinctive: line 19
If installation fails, report the failure instead of changing the source or scope.
≈ “Common” is an estimate: at least 1 in 5 of the 3,592 llms.txt 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
# Skillet > Build agent skills from a reviewable spec, with optional evals for selected behavior. Skillet keeps the required spec and agent instructions reviewable. Eval cases are a separate opt-in for scenarios worth measuring; ordinary authoring does not create or run them. - [Documentation](/) - [GitHub](https://github.com/getsentry/skillet) - [npm](https://www.npmjs.com/package/@sentry/skillet) ## Install the Authoring Skill When asked to install Skillet globally for an agent: 1. Install `skills/skillet-authoring` from `https://github.com/getsentry/skillet` in user scope. 2. Prefer `npx -y @sentry/dotagents@latest --user add getsentry/skillet skillet-authoring`. 3. Do not create project-local skill files or modify the current repository. 4. After installation, run Skillet commands through `npx -y @sentry/skillet@latest` or `pnpx @sentry/skillet@latest`. 5. If installation fails, report the failure instead of changing the source or scope. The latest `skillet status` checks the standard user-scoped installation for a stale authoring contract. When stale, follow `status.next`: reinstall through its scoped dotagents command or the same method that originally installed it, then rerun status before continuing. To continue in the same session, read the reinstalled SKILL.md first; otherwise start a new session. Skillet never silently rewrites user configuration. Custom installation locations require manual reinstallation through their original method. ## Start Here - [Overview](/index.md) - [Quickstart](/quickstart.md) - [Create Your First Skill](/first-skill.md) - [Adopt an Existing Skill](/existing-skill.md) - [Examples](/examples.md) ## Build a Skill - [Specifications](/concepts/specifications.md) - [Write Agent Instructions](/guides/write-agent-instructions.md) - [Write Honest Evals](/guides/write-honest-evals.md) - [Understand Eval Results](/concepts/evaluations-and-lift.md) ## How Skillet Works - [Artifact Lifecycle](/concepts/artifact-lifecycle.md) - [Configure Harnesses](/guides/configure-harnesses.md) - [Sandbox and CI](/guides/sandbox-and-ci.md) ## Reference - [CLI](/reference/cli.md) - [Eval Case YAML](/reference/eval-case.md) - [.skillet.yaml](/reference/configuration.md)
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
Reviews can't be read right now.
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.