Skip to content
Back to skills

Agent Doc Discipline

ASecurity

Writing-time discipline for documents agents consume (agent-facing docs, specs, tickets, .dsh/skills/) — every rule checkable and carrying a why, steps before reference, one meaning in one home, no restating what the environment already says

  • 8 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added October 13, 2026
ai-agentsgorails

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned October 13, 2026

npx -y skills add 1225Sakura/oh-my-deepseekharness --skill agent-doc-discipline --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Agent Doc Discipline?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Agent Doc Discipline
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/1225sakura-agent-doc-discipline/badge)](https://www.skillsdirectory.com/skills/1225sakura-agent-doc-discipline)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: agent-doc-discipline
description: Writing-time discipline for documents agents consume (agent-facing docs, specs, tickets, .dsh/skills/) — every rule checkable and carrying a why, steps before reference, one meaning in one home, no restating what the environment already says
when-to-use: Creating or editing any document an agent reads to act — AGENTS.md / CLAUDE.md, specs, plans, tickets, skill files under `.dsh/skills/`. The companion of minimal-code-discipline: that one disciplines code, this one disciplines papers.
---

# Agent Doc Discipline

Use this skill to apply a writing-time discipline while creating or editing any document an agent consumes to act: agent-facing docs (`AGENTS.md`, `CLAUDE.md`, `CONTEXT.md`, `docs/standards/`), specs, tickets, and skill files (`.dsh/skills/`). The test a document must pass: **a fresh agent session can act on it by reading alone.**

Note: in OMC this discipline was mandatory inside the shipyard flows (drydock seed generation, launch C5 sediment pass). drydock/launch are ported to omd (see the same-named skills), and this discipline is likewise mandatory in their seed generation and C5 sediment pass; other agent-facing doc edits apply it on demand — but any edit to an agent-facing document should survive these rules. Likewise, OMC's `.omc/skills/` learned-skill surface maps to the dsh skill roots (`.dsh/skills/` project, `~/.dsh/skills/` user).

## When Not to Use

Documents written for humans only: `docs/business/` narrative, ADR decision stories, READMEs for human onboarding. They may still follow the rules where it costs nothing, but they are not held to this discipline.

## The Discipline

**Write for a reader with no chat history.** The document is self-describing: nothing leans on "as discussed", a prior session, or knowledge that lives in someone's head. _Test:_ a brand-new agent session can act on the document by reading alone.

**Every rule checkable and carrying a why.** No "keep it clean" — write the observable condition and the reason. _Test:_ an agent can decide pass/fail from the text alone, and the why is stated next to the rule.

**One meaning, one home.** Each rule or definition has a single authoritative place; duplication is drift waiting to happen. _Test:_ changing a behavior is a one-place edit.

**Never restate what the environment confesses.** Package scripts, directory layouts, `--help` output are lookups, not documents — a copied lookup goes stale. Document only what cannot be found by looking: the unwritten convention, the reason behind a choice, the gotcha. _Test:_ every line survives "can this be looked up elsewhere?".

**Steps first, reference behind, rare material behind a pointer.** What to do, in order, at the top; consult-on-demand rules below; low-frequency material pushed behind a pointer whose wording names the branches that should trigger reaching it. _Test:_ the next action is findable within the first screen.

**Every step ends on a completion criterion.** Done must be tellable from not-done — a vague bound invites premature completion. _Test:_ for each step, "how do I know this is finished?" has an answer in the text.

**Prompt the positive.** State the target behavior; a prohibition is reserved for hard guardrails and is always paired with the positive target. _Test:_ every prohibition in the document names the thing to do instead.

**Scrape barnacles on write.** When the document contains stale or redundant material, remove it in the same edit; when it does not, add only the required material and do not invent deletions. A sentence the model already obeys by default pays load for nothing — delete the whole sentence. _Test:_ every line re-read earns its place against "does this change behavior versus the default?", and any stale or redundant material found during the edit is gone.

## Verification

Before reporting a document change done, confirm:

- a fresh session could act on it without asking a human anything
- every rule is checkable and states its why; no rule restates a lookup
- meanings live in one place; pointers name their trigger branches
- stale or redundant material found during the edit was removed, while valid unrelated content was preserved

## State Contract (状态契约)

Agent-doc-discipline **holds no mode state**. It is a writing-time discipline: no `state_write`/`state_clear`, no files under `.omd/`. Its deliverable is the edited document itself.

Files in this skill

  • SKILL.md4.4 KB
  • SKILL.zh.md3.9 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.