{
  "markdown": "![skills cover](assets/gh-repo-cover.png)\n\n# skills\n\nBring your Entire context to your agents with cross-agent skills.\n\n## Why This Exists\n\nThe [Entire CLI](https://github.com/entireio/cli) captures the context behind your\ncode changes: prompts, transcripts,\n[Checkpoints](https://docs.entire.io/cli/checkpoints), and the decisions that led\nto each change, alongside your git history. This repository packages\nagent-invokable workflows that teach coding agents how to use that context across\ndevelopment environments.\n\nInstead of looking up Entire commands yourself, you can ask in plain language and\nlet your agent search prior work, inspect provenance, hand off session state, or\nturn repeated workflows into new skills.\n\n## Install\n\nInstall every skill with the [skills](https://skills.sh/) CLI:\n\n```bash\nnpx skills add https://github.com/entireio/skills --all\n```\n\nInstall one skill:\n\n```bash\nnpx skills add https://github.com/entireio/skills --skill search\n```\n\nSee [Agent-Specific Installation](#agent-specific-installation) for setup\ninstructions by agent.\n\n## Quick Start\n\n> [!NOTE]\n> These skills are most useful in codebases with real\n> [Checkpoints](https://docs.entire.io/cli/checkpoints) and session history. The\n> richer the history, the more context your agent has to work with; in a new or\n> lightly captured repository, some workflows may have less to find or explain.\n\nAfter installing, ask your agent for the workflow you want:\n\n```text\nsearch past work for rate limiting\n```\n\n```text\nexplain src/auth.ts\n```\n\n```text\nwhat happened here: src/auth.ts:42-57\n```\n\n```text\nhand off this session\n```\n\n```text\nturn my release notes workflow into a skill\n```\n\n```text\nreview this branch before merging\n```\n\nFor a guided walkthrough, see the\n[skills tutorial](https://docs.entire.io/skills/tutorial).\n\n## What skills help agents do\n\n| Goal | Example prompt |\n| --- | --- |\n| Research or investigate a topic from recorded history | `research how checkpoint pushing works` / `investigate the flaky login redirect` |\n| Find prior work before making changes | `search past work for the migration` |\n| Understand the intent behind a function, file, or line | `explain parseConfig` |\n| Investigate the latest change to a specific block | `what happened at src/auth.ts:42` |\n| Review branch changes with intent context | `review this branch before merging` |\n| Pick up another agent's work | `hand off the codex session` |\n| Convert repeated work into a reusable workflow | `make a skill from this session` |\n\n## Make search the default research step\n\nAgents pick skills by matching the request against each skill's description, and\ngeneric verbs like \"research\", \"investigate\", or \"look into\" compete with their\nbuilt-in grep and file-reading tools. The `search` and `using-entire` skill\ndescriptions claim those verbs explicitly and ask for the skill as the first\naction, before any file orientation, so a plain \"investigate X\" routes to\n`entire search` first. This is measured to be sufficient on its own (see\n`evals/README.md`).\n\nUser instructions outrank skill descriptions in every agent, so if you want a\nguarantee, add a line to your `CLAUDE.md` or `AGENTS.md`:\n\n```markdown\nWhen asked to research, investigate, look into, or dig into anything in this\ncodebase, run `entire search \"<topic>\" --json --compact --limit 5` first and\nsummarize the prior work before reading source files or grepping.\n```\n\nThe `evals/` directory holds `claude plugin eval` cases that check this routing;\nsee `evals/README.md`.\n\n## Included skills\n\nThe current repository includes these skills. Each skill lives in\n`skills/<skill-name>/SKILL.md`.\n\n### `search`\n\nFinds prior work in your Entire history by topic, repo, branch, author, or time\nwindow, so your agent can bring past context into the current task before making\nchanges.\n\nhttps://github.com/user-attachments/assets/52e50eae-ba53-4f31-ad2e-2a29a767a34f\n\n### `explain`\n\nLooks up the session behind a function, file, or line so your agent can explain\nthe requirement, decision, or original problem that shaped it.\n\nhttps://github.com/user-attachments/assets/ab2e88f6-9bce-417b-ba71-6e93ecad370b\n\n### `what-happened`\n\nStarts from an exact file line, range, or pasted snippet and traces the latest\nchange with `git blame` and Checkpoint context. Useful when reviewing a concrete\nblock, debugging a regression, or asking why that block changed.\n\nhttps://github.com/user-attachments/assets/47db0daa-60bd-4000-b956-2d0e4d980b9b\n\n### `session-handoff`\n\nReads saved or active session context so another agent can pick up the task\nstate, important discoveries, blockers, and next steps without making you\nreconstruct everything manually.\n\nhttps://github.com/user-attachments/assets/0df3b5cd-fe37-4145-af48-138642ccc8bc\n\n### `review`\n\nReviews code changes on the current branch by reading checkpoint transcripts to\nunderstand developer intent, then auditing the diff for issues. Produces\nintent-aware findings with severity levels (`Critical` / `High` / `Medium` /\n`Low`) compatible with `entire review --fix`. Includes a separate\n`references/review-rules.md` for the audit checklist.\n\n### `using-entire`\n\nOrchestrator skill for codebase exploration. Routes user intent to the right\nsub-skill (search, explain, what-happened, review, session-handoff,\nsession-to-skill, session-crosslink) or runs a general exploration flow that\nreads checkpoint history before inferring from code.\n\n### `session-crosslink`\n\nLinks an agent session that ran outside the repo whose commits should record it\n— launched from a higher-level folder, a non-Entire repo, or one repo but\nediting another — to each affected Entire-enabled repo's HEAD commit. Previews\nwith `--dry-run` then amends on confirmation, no manual `cd` orchestration.\n\n## Requirements\n\nThese skills are designed for repositories where Entire has captured useful\nhistory. Some workflows need:\n\n- the [Entire CLI](https://github.com/entireio/cli) installed\n- a git repository with Entire sessions, Checkpoints, or Checkpoint-backed commits\n- [`entire login`](https://docs.entire.io/cli/commands#login) for workflows that\n  search indexed history\n- a GitHub `origin` and pushed, indexed Checkpoints for remote search results\n\nIndividual skills handle missing context differently and should explain what is\nmissing before falling back or stopping.\n\n## Agent-Specific Installation\n\n<!-- prettier-ignore-start -->\n\n<details>\n<summary>Claude Code</summary>\n\n```bash\n/plugin marketplace add entireio/skills\n/plugin install entire\n```\n\n</details>\n\n<details>\n<summary>Codex (OpenAI)</summary>\n\nClone into the cross-client discovery path:\n\n```bash\ngit clone https://github.com/entireio/skills.git ~/.agents/skills/entire\n```\n\nCodex auto-discovers skills from `~/.agents/skills/` and `.agents/skills/`.\nUpdate with:\n\n```bash\ncd ~/.agents/skills/entire && git pull\n```\n\n</details>\n\n<details>\n<summary>Cursor</summary>\n\nCopy skills into the cross-client discovery directory:\n\n```bash\ngit clone https://github.com/entireio/skills.git ~/.cursor/skills/entire\n```\n\nCursor auto-discovers skills from `.agents/skills/` and `.cursor/skills/`.\n\n</details>\n\n<details>\n<summary>Copilot</summary>\n\n```bash\n/plugin install https://github.com/entireio/skills\n# or\ngit clone https://github.com/entireio/skills.git ~/.copilot/skills/entire\n```\n\nCopilot auto-discovers skills from `.copilot/skills/`.\n\n</details>\n\n<details>\n<summary>Gemini CLI</summary>\n\n```bash\ngemini extensions install https://github.com/entireio/skills\n```\n\nUpdate with:\n\n```bash\ngemini extensions update entire\n```\n\n</details>\n\n<details>\n<summary>OpenCode</summary>\n\nCopy skills into the cross-client discovery directory:\n\n```bash\ngit clone https://github.com/entireio/skills.git ~/.agents/skills/entire\n```\n\nOpenCode auto-discovers skills from `.agents/skills/`, `.opencode/skills/`, and\n`.claude/skills/`.\n\n</details>\n\n<!-- prettier-ignore-end -->\n\n## Contributing\n\nSee [CONTRIBUTING.md](CONTRIBUTING.md) for local development, validation, and pull\nrequest guidance. Please also follow the\n[Code of Conduct](https://github.com/entireio/.github/blob/main/CODE_OF_CONDUCT.md).\n\n## Support\n\nFor bugs, feature requests, or questions, open a\n[GitHub issue](https://github.com/entireio/skills/issues). You can also join the\n[Entire Community Discord Server](https://discord.gg/jZJs3Tue4S).\n\n## License\n\nThis project is licensed under the terms in [LICENSE](LICENSE).\n",
  "bytes": 8391,
  "sha": "ad2e600034247a08f6b97031476ce9ffb7ea851903e0e09f3eeef10a6ec3ba03",
  "repo_slug": "entireio/skills",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_entireio_skills_a0b9e598/readme"
}