
indexion-readme
README construction — initialize template structure, generate per-package READMEs from doc comments, plan writing tasks, assemble root README from docs/ and package READMEs via doc.json config, and verify edits with plan drift.
- Installs
- 5.1K
- Last 24h
- —
- Trending rank
- —
- Security
- Pass
Install
Works with every agent the skills CLI supports.
npx skills add https://github.com/trkbt10/indexion-skills --skill indexion-readme- 01
Install
Run the command in your project. The CLI asks which agents to install for.
- 02
Describe the task
Ask your agent for something this skill covers. It loads indexion-readme automatically when relevant.
- 03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.
Security audits
Independent scans of the skill's instructions and bundled files.
SKILL.md
The instructions your agent reads when the skill is triggered.
indexion readme — README Construction
Build project READMEs from templates, doc comments, hand-written prose, and
per-package READMEs. This skill covers the construction side: scaffolding,
generating, planning, assembling, and verifying. For evaluating existing
documentation, see indexion-documentation.
Where things live
The conventions vary by project; check what actually exists before editing:
| Asset | Location patterns seen |
|---|---|
| Config | doc.json (repo root) or .indexion/readme/doc.json |
| Per-package template | docs/templates/readme.md (no SoT constant; declared via --template) |
| Static prose | docs/intro.md, docs/installation.md, docs/license.md, … |
| Per-package READMEs | cmd/<name>/README.md, src/<name>/README.md |
| Assembled root README | usually README.md. Some projects use a .mbt.md suffix so the file is also a MoonBit doctest module — when so, README.md is a symlink to README.mbt.md. |
.indexion.toml | [doc] config_path / per_package auto-load doc.json |
The first thing to check is git diff / ls -la for: a symlink on README.md, a doc.json (at root or under .indexion/readme/), and .indexion.toml. The presence of these tells you whether the README is hand-maintained, build-assembled, or a hybrid.
Workflow overview
doc init → .indexion/readme/template.md + doc.json (greenfield)
edit doc.json + docs/*.md → declare sources of truth
doc readme --per-package → cmd/<pkg>/README.md (API skeletons; non-overwriting)
hand-write each per-package README's Overview / Usage / Options / Examples
doc readme --config → assembled root README
plan drift <prev> <new> → verify the assembled output (or hand-edit) is purely additive
Step 1: Initialize (greenfield only)
indexion doc init <project-dir>
Creates .indexion/readme/template.md + .indexion/readme/doc.json. Skip this step on a project that already has doc.json or .indexion.toml pointing at one.
Step 2: Configure doc.json
{
"$schema": "./schemas/doc-config.schema.json",
"version": "1.0",
"spec": "moonbit",
"output": { "format": "markdown", "filename": "README.md" },
"packages": [
{
"path": "cmd/<name>",
"title": "<command name>",
"include_in_root": true,
"sections": ["overview", "usage"]
}
// …one entry per package that should appear in the assembled root README
],
"root": {
"output": "README.md",
"sections": [
{ "type": "static", "file": "docs/intro.md" },
{ "type": "toc", "title": "Commands" },
{ "type": "packages", "filter": "cmd/**" },
{ "type": "static", "file": "docs/installation.md" },
{ "type": "static", "file": "docs/license.md" }
]
}
}
Root section types:
static— include a markdown file verbatimtoc— insert a table-of-contents headingpackages— pull in entries from thepackagesarray, filtered by glob
Package fields:
include_in_root— whether to include in the assembled root READMEsections— README headings to extract; honored by per-package extraction flows. Note that{ "type": "packages" }in the root currently emits a table of package links, not the rich Overview/Usage expansion implied by per-packagesections. See "Known limitation" below.
Step 3: Generate per-package READMEs
indexion doc readme --per-package src/ cmd/
Generates README.md in each package directory that doesn't already have one. Non-overwriting — existing per-package READMEs are left alone.
The skeleton is API-only (extracted from /// doc comments via KGF). Treat it as a starting point and hand-write the prose sections (Overview, Usage, Options, Examples) afterwards.
# Single package to stdout
indexion doc readme src/kgf/lexer/
# Single package to a file
indexion doc readme -o=README.md src/kgf/lexer/
Note on side effects: doc readme --template=<t> <paths...> (the template-based mode below) walks the given paths and auto-creates per-package READMEs for any package that lacks one — even without --per-package. If you run it on a broad path (cmd/, src/), expect new files in unrelated packages. Run with a narrow path or grep git status afterwards to clean up unintended creations.
Step 4: Write static content
Create docs/intro.md, docs/installation.md, etc. — whatever your root.sections references. These are the hand-written prose; the assembler pulls them in verbatim.
Step 5: Generate writing plans (optional)
indexion plan readme --template=docs/templates/readme.md --plans-dir=.indexion/plans src/
Emits per-section writing tasks for manual or LLM-assisted authoring.
Step 6: Assemble the README
# Config-driven (preferred; doc.json controls layout and packages list)
indexion doc readme --config=doc.json
# Template-driven (alternative; {{include:…}} and {{packages}} placeholders)
indexion doc readme --template=docs/templates/readme.md -o=README.md cmd/
The config path can live at repo root or under .indexion/readme/. With .indexion.toml's [doc] config_path = "…", the --config= flag becomes optional.
Step 7: Verify with plan drift
After regeneration or any hand-edit of the root README, verify the change is purely additive (no silent deletions, no reflowing of unrelated sections):
# Snapshot the previous version
git show HEAD:README.md > /tmp/README.before.md
# Compare against the new version
indexion plan drift --top=20 /tmp/README.before.md README.md
What to look for in the output:
Drift terms in /tmp/README.before.md (missing on the other side): (none)— nothing was removedDrift terms in README.md (missing on the other side): …— exactly the new vocabulary you intended to add (command names, new flags, new concepts)Cosine similaritynear 1.0 for a small additive change; substantially lower if you reshaped a section
For CI integration:
indexion plan drift --vocab-threshold=0.05 /tmp/README.before.md README.md
# exits 1 if cosine_distance > 0.05 — useful as a guard against accidental large rewrites
This same workflow applies to translated README pairs (README.md ↔ README-ja.md): cross-lingual drift detection works natively because vocab sub-tokenization delegates to the natural-language KGFs in kgfs/natural/.
Template syntax
The template file supports {{placeholder}} substitution:
| Placeholder | Expansion |
|---|---|
{{include:path}} | Contents of the file (relative to project root) |
{{packages}} | All discovered packages (filtered by CLI --include / --exclude) |
{{module_doc}} | Module-level documentation only |
.indexion.toml integration
[doc]
config_path = "doc.json" # auto-loads doc.json without --config
per_package = true # makes `doc readme <path>` default to --per-package
Explicit --config=… always takes priority.
Known limitation: packages root section produces a table, not rich expansion
The doc-config.schema.json permits sections: ["overview", "usage", …] on each packageEntry, but the current doc readme --config implementation does not expand those sections inline when emitting { "type": "packages" } in the root. The output is a markdown table of package links with empty descriptions.
Two practical consequences:
- If the project's checked-in root README has rich per-command Overview / Usage paragraphs, they were not produced by
doc readme --configas it stands now. They are hand-maintained. Diffdoc readme --config -o=/tmp/regen.mdagainst the checked-in README to see how much of it is hand-curated; very large diffs mean the README is mostly hand-maintained. - For new commands, you currently need to also hand-edit the rich section into the assembled README, in addition to adding the package entry to
doc.jsonand writing the per-package README. Use theplan driftverification above to confirm the hand-edit only adds, never removes.
If you fix this limitation (so { "type": "packages" } honors per-entry sections), update this skill to remove this section.
Common pitfalls
"doc readme --per-package generated nothing"
- All packages already have READMEs. The command only creates new files, never overwrites. Delete existing READMEs first to regenerate.
"doc readme --template … on cmd/ created READMEs in packages I didn't touch"
- Template mode auto-generates missing per-package READMEs as a side effect.
Either pass a narrow path, or revert unintended creations from
git status.
"Auto-generated per-package READMEs are just API listings"
- By design. Hand-write Overview, Usage, Options, Examples. For CLI commands,
the authoritative behavior comes from
indexion <command> --help.
"My hand-edit to README.md will be wiped out by doc readme --config"
- It will if the assembler ever produces the rich shape (see "Known
limitation"). Until then, the assembler produces a strict subset (the
table) and your hand-edits to the rich sections survive. Always run the
plan driftcross-check to be sure.
"README.md edit destroyed unrelated sections"
- Run
plan drift HEAD:README.md README.mdand look at the "missing on the other side" output for the previous version. If it lists anything other than(none), you removed content.
FAQ
What does the indexion-readme skill do?
README construction — initialize template structure, generate per-package READMEs from doc comments, plan writing tasks, assemble root README from docs/ and package READMEs via doc.json config, and verify edits with plan drift.
How do I install indexion-readme?
Run npx skills add https://github.com/trkbt10/indexion-skills --skill indexion-readme in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/indexion-readme/ (personal) or .claude/skills/indexion-readme/ (project).
Which AI agents can use indexion-readme?
Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.
Is indexion-readme safe to install?
Security partners have reviewed it: Gen Agent Trust Hub passed; Socket passed; Snyk passed. As with any skill, read the SKILL.md and bundled scripts before installing.
Who maintains indexion-readme?
It is published by trkbt10 in trkbt10/indexion-skills. It has 5,100 recorded installs.
More from trkbt10/indexion-skills
indexion-sdd
trkbt10/indexion-skills
Generate SDD requirements from RFCs/specs and quantitatively verify implementation conformance. spec draft → spec align → spec verify → automated validation loop with codex/claude. Operate spec-to-impl drift gates as CI checks.
indexion-documentation
trkbt10/indexion-skills
Documentation analysis — assess coverage, detect code-to-doc drift with plan reconcile, visualize dependencies with doc graph. Answers "what needs docs?" and "are docs still accurate?"
indexion-refactor
trkbt10/indexion-skills
After writing code, detect and clean up duplication at three levels — copy-paste blocks, cross-package shared code, unnecessary wrappers, and concept-level SoT violations. Detect with indexion, fix, and verify.
indexion-wiki
trkbt10/indexion-skills
Project wiki lifecycle — create pages, track source changes with ingest, update stale pages, lint structural integrity, detect code-to-doc drift with reconcile, and manage search indexes. A runbook for navigating and maintaining the wiki as an agent.
indexion-segment
trkbt10/indexion-skills
Split text into contextual chunks for RAG/embedding pipelines. Document segmentation and section extraction using window, tfidf, punctuation, or hybrid strategies chosen by intent.
indexion-kgf
trkbt10/indexion-skills
Debug and inspect KGF specs — view tokenization results, parse trees, and extracted edges from source files. Use when adding/fixing language support or when indexion's analysis output looks wrong.
Popular docs & writing skills
grill-with-docs
mattpocock/skills
A relentless interview to sharpen a plan or design, which also creates docs (ADR's and glossary) as we go.
lark-markdown
open.feishu.cn
飞书 Markdown:查看、创建、上传、编辑和比较飞书中的原生 Markdown 文件。当用户要操作飞书 Markdown 文件,或比较其远端版本及本地草稿时使用。纯本地 Markdown 文件操作不触发本 skill。不负责将 Markdown 导入为飞书在线文档,也不负责文件搜索、权限、评论、移动、删除等云空间管理操作。
writing-beats
mattpocock/skills
Writing, exploit; assemble raw material into a journey of beats, grounding each term before a beat leans on it.
writing-shape
mattpocock/skills
Writing, exploit: shape raw material into an article, paragraph by paragraph.
writing-fragments
mattpocock/skills
Writing, explore: mine raw fragments, no structure yet.
writing-for-agents
mattpocock/skills
Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.