readme-status
Content-editor-posture skill that brings a product README's Status section up to what is built.
Skill by eyalgo
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/eyalgo-readme-status
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: writing
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/readme-status
What's inside
- Skill name
readme-status- Description
- Content-editor-posture skill that brings a product README's Status section up to what is built. One run reduces the product's what-left file against the tip of main, removing only the entries whose whole want has merged, then rewrites the Status section from the reduced file, every built sentence citing the commit that brought it in, and makes one commit in the product's repository. Writes no other file and never adds an entry. Halts outside Content Editor posture. Run when a product has merged work and its README Status section may be out of date.
- License
- MIT
- Metadata
- version: 1.2.0
- Instructions
- 130 lines in SKILL.md · view on GitHub
Files (7)
- SKILL.md
- claude.md · intelligence/
- codex.md · intelligence/
- gemini.md · intelligence/
- ollama.md · intelligence/
- launcher.md
- thin.md · prompts/
SKILL.md (first lines)
# readme-status **The prompt owns judgment, and this file owns the loop.** `prompts/thin.md` decides which wants are wholly built, which commit brought each in, and what the Status section says. This file resolves the product, validates both targets, builds both results from original bytes and makes the one commit. It reads that one prompt and no other judgment body. That prompt also carries the published sentence shapes and outcome messages; the posture refusal below runs before it is read. ## When to Use - A product has merged work and its README `## Status` section, or its what-left file, may no longer say what is built. Start it with `<CORE>/launchers/skill/launcher.sh readme-status <project>`. - **Content Editor role only.** Outside a `content-editor` session or subagent, the run changes no file, makes no commit, and its final message says it needs Content Editor posture and to open a session or subagent as `content-editor`. It reads no project input first, the same refusal the seat's other skill gives (`<CORE>/roles/loaders/content_editor_global.md`, *Skills toolbox*). **Refused for posture — the session is not a Content Editor session** > readme-status changed no file and made no commit. > It needs Content Editor posture. > Open a session or subagent as `content-editor` and run readme-status again. **Do not use for:** - Adding, rewording or reordering a what-left entry. An entry is added by whoever changes the Agari or PRD (`<CORE>/doctrine/setup/kensaiapp_structure.md` § 3, *The what-left file*). - Editing a README outside its Status section, or creating a README. - The words of one change. That is `development-content`'s, with its own inputs and output. - Writing or amending a standing wording rule. A rule the Status section needs and the standard lacks is a finding in the final message, for the Content Manager. ## Input **The project** — `<project>`, one name. The change folder, a change key and a halt report are not inputs and the run never reads them. ## Workflow Run from the instance root. After the posture gate, read `prompts/thin.md` for the judgment and the outcome sentences. A pre-write refusal changes no file and makes no commit; its final message opens with the refusal sentence and names the condition that held. A commit-builder refusal uses the separate message that says the files were written and remain uncommitted. ### 1 — Resolve the destination Read `system/projects/<project>/project.md`. Refuse if there is no card, or no `Product path:` line. Resolve `Product path:` as the launcher does: relative against the instance, absolute directly, `~` expanded. Refuse if no directory is there. Find the repository holding it with `git -C <product> rev-parse --show-toplevel`, and refuse if there is none. That checkout must already be on `main`. Never switch a branch and never open a change worktree. A product in another repository commits there, and an in-repo product commits in the containing instance repository. ### 2 — Read and validate, before any write Read the what-left file at `<product>/docs/what_left.md`, to the grammar and the readings of `<CORE>/doctrine/setup/kensaiapp_structure.md` § 3 (`:2301-2307`), and `<product>/README.md`. Refuse for each of these, one condition at a time: - the what-left file does not exist (never given), or holds neither an entry nor the empty marker, or holds both; - an entry lacks a `**Wanted:**`, `**Source:**` or `**Missing at:**` line, or its `Missing at` commit is not carried by `main`; - the README does not exist, or does not have exactly one line reading `## Status`; - either target has uncommitted changes, or an unrelated path is staged; - `main` is unavailable, or any other condition that makes either rewrite impossible. Record the original bytes of both targets, `main`'s tip and the exact start and end of the Status body. Read the product's `docs/CONTENT.md`, or `system/doctrine/writing_bar.md` when it has none. ### 3 — Judge Evidence a product check needs runs in a **disposable copy** — a fresh clone at the commit under test (`git clone --quiet <repo> <scratch> && git -C <scratch> checkout --quiet <sha>`; never `git archive`, which drops the history a check may read), never in the target checkout (`<CORE>/roles/CROSS_ROLE.md` § *Every seat — a run leaves the product's main checkout as it found it*). Run the checks from a subshell, never `cd` your own shell into the copy, and remove the copy when you have read the result, in a command of its own run from outside it. Apply `prompts/thin.md` to the judgment with the retained entries, the evidence, the original Status text and the standard. A refusal it returns for a built want with no establishable commit is acted on here, before any write. ### 4 — Build both candidates, then check them Prepare one candidate pair plus the transient evidence before any target write. - The what-left result removes only the selected original three-line entry spans. The H1 and lead sentence stay byte-identical. When the last entry goes, the file holds the line `Nothing is left to build.` alone. The file is never deleted. - The README result replaces only the Status body: from after the `## Status` line to before the next line starting `## `, or the end of the file. - Compare the unchanged bytes outside the two edits. A zero-length old hunk range is an insertion point after the heading and before the next heading, including an empty Status body. - Check that every retained want is represented as not built, that each built sentence carries its own nine-hexadecimal commit in an inline code span, with `git merge-base --is-ancestor <sha> main` true for each, and that no term the standard rejects is used. ### 5 — Re-check, then commit once Immediately before the first write, re-check `main`'s tip, both original byte sequences, the targets' cleanliness and the staged set. Refuse on a changed input, a moved `main`, or an unrelated staged path. If both candidates are byte-identical to their inputs, make no commit and report that nothing changed. Otherwise write only the changed targets and make **exactly one commit** through `<CORE>/tools/commit_builder.py`, following `<CORE>/doctrine/governance/commit_via_builder.md`: `--repo` is the product's repository, the staged files are the changed targets only, **`--actor content-editor`**, **`--skill readme-status`**, the runtime's surface and identity, and `--reasoning-effort` when the launch resolved one. The directional line follows `<CORE>/doctrine/governance/commit_format.md` § Directional trailer. The body carries one line per removed entry, `Merged in <commit>: <the entry's Wanted text, verbatim>`; write the body through a file, because Wanted text may hold backticks and shell metacharacters. When the builder refuses, say so in the final message, with its message verbatim. Use no empty commit, no index waiver and no second commit to get past it. ## Final message The one outcome (changed, nothing changed, or a refusal), every built claim left out because no commit could be established, and any wording rule the standard lacks, in the sentences of `prompts/thin.md`. All findings stay in the final message. The run writes no report file, writes nothing in a satellite and writes no file but the two targets.
From eyalgo/kensaiapp, MIT.
More from eyalgo/kensaiapp
- 0.570score
development-design
eyalgo
Architect-posture skill for ongoing work on an existing project.
SkillAFreshness: AClaude Code +9 - 0.570score
development-content
eyalgo
Content-editor-posture skill for ongoing work.
SkillAFreshness: AClaude Code +9 - 0.565score
first-what-left
eyalgo
CPO-posture skill that writes a product's first what-left file, docs/what_left.md, once the project's Agari is settled.
SkillAFreshness: AClaude Code +9 - 0.565score
product-prd
Eyal Golan
CPO-posture skill that authors a project's PRD — the product source of truth and the head of the definition sequence.
SkillAFreshness: AClaude Code +9 - 0.565score
product-srs
Eyal Golan
CTO-posture skill that authors a project's SRS — the technical source of truth, written off the settled PRD alongside the CDO's UXD.
SkillAFreshness: AClaude Code +9 - 0.565score
refactor-for-optimize
eyalgo
Refactor code for performance — algorithmic complexity, data structure selection, unnecessary I/O, hot-path improvement, and memory usage.
SkillAFreshness: AClaude Code +9 - 0.565score
refactor-for-clarity
eyalgo
Refactor code for clarity without changing behavior.
SkillAFreshness: AClaude Code +9 - 0.561score
refactor-for-reuse
eyalgo
Refactor changed code to eliminate duplication, extract shared logic, and remove dead code.
SkillAFreshness: AClaude Code +9 - 0.560score
product-uxd
Eyal Golan
CDO-posture skill that authors a project's UXD — the experience source of truth, written off the settled PRD alongside the CTO's SRS and read by the SDD.
SkillAFreshness: AClaude Code +9 - 0.560score
product-content
Eyal Golan
Content-Manager-posture skill that authors a project's verbal language (CONTENT.md) — the standing source of truth for everything a product says to a person…
SkillAFreshness: AClaude Code +9 - 0.560score
experience-visual
Eyal Golan
CDO-posture skill that authors a kensaiapp instance's visual experience layer — the fence and the direction every product under that instance inherits…
SkillAFreshness: AClaude Code +9 - 0.560score
product-design
Eyal Golan
CDO-posture skill that authors a project's design language — the visual source of truth in Google's DESIGN.md format, written off the settled UXD's…
SkillAFreshness: AClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 19, 2026 (10 days ago)
- Last release
- Oct 22, 2026 (7 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- Version
- 1.2.0
- License
- MIT
- Repository
- github.com/eyalgo/kensaiapp
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from eyalgo
- 0.570score
development-design
eyalgo
Architect-posture skill for ongoing work on an existing project.
SkillAFreshness: AClaude Code +9 - 0.570score
development-content
eyalgo
Content-editor-posture skill for ongoing work.
SkillAFreshness: AClaude Code +9 - 0.565score
refactor-for-optimize
eyalgo
Refactor code for performance — algorithmic complexity, data structure selection, unnecessary I/O, hot-path improvement, and memory usage.
SkillAFreshness: AClaude Code +9 - 0.565score
refactor-for-clarity
eyalgo
Refactor code for clarity without changing behavior.
SkillAFreshness: AClaude Code +9 - 0.565score
first-what-left
eyalgo
CPO-posture skill that writes a product's first what-left file, docs/what_left.md, once the project's Agari is settled.
SkillAFreshness: AClaude Code +9 - 0.561score
refactor-for-reuse
eyalgo
Refactor changed code to eliminate duplication, extract shared logic, and remove dead code.
SkillAFreshness: AClaude Code +9
Cite this page
readme-status skill. Agent Mart. https://agentmart.dev/skills/eyalgo-readme-status (last updated 2026-10-24)