Skill Seekers Bootstrap
Self-document Skill Seekers as an installable Claude Code skill.
The documentation makes the local-codebase analysis, output locations, and optional Claude Code installation reasonably transparent. However, the script is described as removing the output directory without confirmation, installation can write to an agent directory, and rollback, privilege isolation, and sensitive-data handling are not specified. No malware or credential-theft red line is evidenced.
Commands, output structure, validation steps, and troubleshooting are fairly detailed, and unit and end-to-end tests are claimed. Static review provides no test files or CI results, while command, directory, and version descriptions contain inconsistencies, so the score remains below the static ceiling.
The intended directory input, output directory, workflow, and principal commands are clear. However, this path documents self-documentation of Skill Seekers, while the supplied object description also claims website, GitHub, and PDF conversion; non-fit boundaries, trigger conditions, Chinese support, and mainland-China reachability are not documented.
The document has good layering, examples, troubleshooting, version information, changelog material, and supplied MIT license metadata. It lacks a clearly named maintainer and formal update path, and its Last Updated date conflicts with the changelog date, so full marks are not justified.
The core workflow plausibly produces and installs a Claude skill from a codebase, with outputs that appear directly usable after review. Nevertheless, output correctness, coverage, and quality are asserted rather than verified for this revision, and manual review remains necessary.
The files provide commands, named tests, metrics, and examples as audit clues. They do not include the test implementations, CI execution evidence, or independent corroboration in the supplied material, and the production metrics cannot be verified statically.
- Use a dedicated output directory because the documented script removes that directory without confirmation.
- The claimed tests, coverage, and production metrics are not independently verified in the supplied material; review generated content before installing it into Claude Code.
- This path primarily covers local codebase self-documentation; do not assume website, GitHub, or PDF conversion capability, Chinese support, or mainland-China reachability from it.
- Note the date inconsistency: Last Updated is 2026-02-18, while the v3.6.0 changelog is dated 2026-05-30.
What it does & when to use it
Bootstrap Skill uses Skill Seekers to analyze the Skill Seekers codebase itself and generate a Claude Code skill containing documentation, CLI references, API documentation, code patterns, and test examples. It combines a handcrafted header with generated analysis and validates the resulting structure and YAML fields. The output can be installed into Claude Code or regenerated in CI/CD. It is best suited to developers who need an operational reference for Skill Seekers or a maintainable self-hosting workflow.
Runs ./scripts/bootstrap_skill.sh or skill-seekers create --directory . against the local Skill Seekers repository; analyzes Python source, tests, configuration, and documentation; detects design and architectural patterns; extracts test examples, configuration patterns, API references, dependency graphs, and how-to guides; merges generated content with scripts/skill_header.md; validates frontmatter, required fields, and structure; creates output/skill-seekers/ with SKILL.md, references, and metadata; and can install the result into Claude Code.
- A new Skill Seekers user wants Claude Code to explain the project's CLI, APIs, and common workflows.
- A maintainer wants to regenerate project documentation after code changes, including updated patterns and test examples.
- A team wants to generate and preserve or publish the self-hosting skill on a recurring CI/CD schedule.
- A developer needs source-derived design-pattern, architecture, API, or programmatic-usage references.
Pros & cons
- Covers source code, tests, configuration, and documentation rather than producing only a high-level overview.
- Generates design- and architectural-pattern material, API references, dependency graphs, and how-to guides.
- Validates frontmatter, required fields, YAML syntax, and content structure.
- Supports both local execution and recurring CI/CD generation.
- The skill is specifically designed to document Skill Seekers itself; the source does not establish that it automatically documents arbitrary projects.
- The end-to-end workflow is described as taking about 2–5 minutes, with some C3.x analysis stages taking longer.
- Generated content depends on the current state of the local source, tests, configuration, and documentation.
- The supplied sources contain inconsistent version metadata: the skill states 3.6.0 while the README identifies the repository as 3.7.0; its update and changelog dates also differ.
How to install
Install the collection with pip install skill-seekers. From the Skill_Seekers repository root, run ./scripts/bootstrap_skill.sh. After generation, run skill-seekers install-agent output/skill-seekers --agent claude, or use the README's cp -r output/skill-seekers ~/.claude/skills/ command to install it for Claude Code.
How to use
From the repository root, run ./scripts/bootstrap_skill.sh. For the manual workflow, run skill-seekers create --directory . --output output/skill-seekers --name skill-seekers, merge scripts/skill_header.md with the generated file, then run skill-seekers install-agent output/skill-seekers --agent claude. The source also shows programmatic usage through scrape_codebase(...) followed by install_to_agent(...).