charter
Write, revise or review a charter: the one document that tells an executor what to deliver, what is out of bounds, what the owner already decided, and how each success check will be proven. Use when work needs a written brief before someone else does it, or when asked to write, change or check a charter.
- License 许可
- Apache-2.0
- 最近更新
- 2026-11-14
- Stars
- ★ 1.6k
- Forks
- 490
01 前言 Front Matter
- name
- charter
- requires
- { "harness": [ { "need": "required", "slot": "4.shell" }, { "need": "required", "slot": "4.file-read-write" } ], "programs": [] }
- description
- Write, revise or review a charter: the one document that tells an executor what to deliver, what is out of bounds, what the owner already decided, and how each success check will be proven. Use when work needs a written brief before someone else does it, or when asked to write, change or check a charter.
02 资源文件 3
- 参考文档references/questions.md847 B
- 参考文档references/template.md2.4 KB
- 脚本scripts/selftest/run-selftest.sh1.5 KB
03 SKILL.md 正文
charter
A charter is one markdown file that says what to deliver and how everyone will know it was
delivered. A task is named by its charter
filename. charter writes it; deliver carries a released charter out. Any kind of project
uses it: software, documents, campaigns, research, operations. The owner is whoever owns the
work (often the user); speak to them in the language they use.
Prerequisites: none to use it; its selftest needs bash and python3.
Read references/template.md for the file layout and references/questions.md before asking
the owner anything.
Rules that keep a charter useful
- Preserve the owner's words and decisions. Never invent a decision, a fact or a requirement. A missing choice that would change the outcome is asked once, one question at a time, at the end of a message; independent work continues while it waits, and silence is not approval.
- Read the existing work first (files, documents, data, earlier decisions). Ask nothing they already answer.
- Every success check names what will be observed, where, and how, and can be answered pass or fail by a real run or inspection. The expected result comes from the requirement or from independent data, never from pasting the program's own output. For a defect, the check reproduces the defect first.
- Before adding a precondition, check, section or approval step, name the failure it detects and what would change if it failed. If nothing changes, leave it out. By default no checksums, sign-off records, warm-up runs, baselines, fixed snapshots or report formats are entry conditions.
- By default a security check, a full test suite or a full evaluation enters a charter only if the owner names it. Propose it as a question that says what it would check and what it would block; do not write it in as a requirement on your own judgment. A project may choose otherwise.
- Say what happens on failure: which operation is logged with its error, what continues. A record or report failure never disables the main result; a real failure stays a failure.
- Time is finite. A check that takes long states its expected duration from existing timings; by default no full suites unless the owner asks.
- State what the executor may not do without the owner: publish, send to other people, spend, delete, overturn a recorded decision.
Write
- Read the request, adjacent notes and the existing work. Cite settled decisions without copying or locking their source. When the work depends on a third-party product, service or version, read its current official documentation and record the facts that constrain the work with the page and version.
- Draft Owner decisions, Outcome, Scope and non-goals, and Success checks from the template. Length follows the work: a few lines for a small task, a sliced charter (named slices, each with its own checks) for a big one. There is one kind of charter, not sizes.
- Read it as the executor: for each check, could someone tell pass from fail without asking? Walk one real example through it. Fix the gaps that would make the result fail. Do not run the future work to plan it.
- One independent review for the owner's intent and concrete defects (
peer-reviewwhen it is installed, otherwise a second agent). Fix material findings; no second round unless the owner asks. - Tell the owner what the charter will deliver and any genuine open decision, in plain
words, ending with the one question. After the owner approves, set
status: releasedand hand the path to the executor, or to whoever supervises the work.
By default an agent alone never releases a charter: it records the owner's decisions. When the
owner is away, leave the charter as a draft for them to release; a project may choose
otherwise.
Revise
A scope change updates the affected decisions and checks in the same edit, sets updated, and
leaves the rest alone. A replaced charter gets status: superseded and a pointer to its
successor. Delivery records delivered only through deliver; never infer it from the header.
Review only
Read the charter and the affected work. Report concrete mismatches, their consequences and the smallest corrections in one local file; do not edit the reviewed charter. Questions go in the conversation. Applying accepted corrections is Revise.
04 同仓库更多 Skills
migrate-to-e2e
Moves an existing Playwright, Cypress, Selenium, Appium, Detox, Maestro, or TesterArmy suite to e2e one slice at a time, with the old suite running beside the new one until each ported test is proven. Keeps a MIGRATION.md progress file so any later session resumes where the last stopped. Optionally plants a defect to show both suites catch it. Use for /migrate-to-e2e, "migrate our tests to e2e", "port these Maestro flows", "move our TesterArmy tests", or when a project has an e2e config and an older suite still running.
e2e
Writes, runs, and debugs end-to-end tests for web and mobile apps with e2e, the agentic test runner, which mixes agent.act goals with exact locator assertions on a Playwright browser or an iOS or Android device. Also drives a project's running app live over the e2e MCP server to look at a screen, check a UI change, take a screenshot or record a video, and try a locator; and runs bug bashes, parallel e2e explore runs with each bug proven by a repro test. Use when a project depends on e2e or has an e2e.config.ts, when asked for end-to-end, browser, mobile, or agentic UI tests, to check a change in the running app, to QA or bug bash an app or branch, or when an e2e run fails.
create-verification-skill
Generate a project-local verify-<app> skill on top of e2e, so any coding agent can launch the app, check it is worth driving, exercise a feature the way a user does, keep evidence, and bug bash it. Writes the skill, a feature map with one explore charter per feature, and proves it once. Use for /create-verification-skill, "make a verification skill for this repo", or when a project has no scripted way to prove UI behavior.
deja-search
deja-vu memory — search the user's past AI coding sessions with the deja CLI. Holds what they already settled and never wrote into the repo (naming rules, banned or already-tried tools, required checks before merge, preferred values, who owns what). Use before your first edit in a task, before you call a change done or ready to merge, before you recommend a tool, library, design or reviewer, before re-debugging an error that may already be solved, when they say 'didn't we fix this before', 'what did we decide about X' or 'I already have that', and before telling them something on this machine does not exist.
deja-reflect
End-of-session reflection and memory curation with deja-vu. Review what was decided, what commands broke and what repaired them, and record durable project rules or decisions with deja remember so the next session starts with them.
revid
Make, edit, export and post AI videos with Revid (revid.ai) through the Revid MCP server. Use whenever the user wants a short-form or long-form video made (from an idea, a script, a URL, an article, a PDF, a tweet, a product page, their own clips, or a song), wants to find what is going viral on TikTok, YouTube, Instagram or X before making something, wants to edit or re-export an existing Revid project, generate a standalone AI image or clip, make a series episode, or publish and schedule a finished video to connected social accounts. Covers setup (hosted MCP with OAuth, or an API key), the 36 tools, credits, polling and the rules for publishing. Routes to the more specific revid-* skills when one fits.