skillZs
★ LIVE SKILL TAGS ★
>>> LIVE SKILLS INDEX <<<
* OPEN SOURCE *
NO LOGIN, NO TRACKING
※ REAL INSTALL DATA ※
← back to all skills
davidlee/doctrine118 installs

canon

Use before making architectural or workflow choices, when correctness depends on project governance, or when entering an unfamiliar subsystem — to load the project's authoritative rules before you act on assumptions.

How do I install this agent skill?

npx skills add https://github.com/davidlee/doctrine --skill canon
view source ↗

Is this agent skill safe to install?

  • Gen Agent Trust Hubpass

    This skill provides a structured method for the AI agent to load and follow project-specific governance and architectural rules. It uses local project documentation and a dedicated CLI tool to ensure the agent acts with project-specific context. No security vulnerabilities or malicious patterns were identified.

  • Socketpass

    No alerts

  • Snykpass

    Risk: LOW · No issues

What does this agent skill do?

Canon

Understand project governance before you act, so you do not spread heresy.

Articles of truth

Identify the authorities that bear on your task, then read the ones that carry meaning for it — not all of them, every time. The CLI is the source of truth — see [[mem.fact.doctrine.cli-source-of-truth]].

  • CLAUDE.md, AGENTS.md — standing instructions and conventions. Project conventions are documented in [[mem.pattern.doctrine.conventions]].
  • project governance - list each of these; read any which are potentially relevant in full, with doctrine <kind> show <id>:
    • doctrine spec list - specs describing product intent / technical architecture
    • doctrine adr list - architecture decision records
    • doctrine policy list - policies
    • doctrine standard list - standards of practice
    • or jump straight to the ones bearing on your task: doctrine search "<keywords>" -k governance. Policies and standards sit outside the default kind scope, so a bare search misses exactly the rules you came for — lib:reference/using-doctrine.md has the scope and the widening flags.
  • the governing slice's design.md — canon for this change's design intent.

For implementation truth (gotchas, patterns, invariants tied to files or commands), /retrieve-memory rather than rediscovering it.

Read entities tier-aware — via show, never a single raw file. The storage rule + read-via-show discipline are resident in the boot digest and detailed in lib:reference/using-doctrine.md. The storage model and storage rule are documented in [[mem.concept.doctrine.storage-model]]; storage tiers: authored, runtime, derived — see [[mem.fact.doctrine.storage-tiers]]. Reading one tier and concluding "empty" is false witness.

The mortal sins

Beware — and refuse to commit:

  • assumption — acting on what you guessed instead of what you confirmed.
  • guesswork — inventing ids, command shapes, or file locations.
  • duplication — a parallel implementation where a seam already exists.
  • architectural infidelity — violating the design or an ADR.
  • heretical coupling — new imports or dependencies that breach boundaries.

If governance is missing, contradictory, or ambiguous on a choice that matters, stop and /consult rather than normalizing around a guess.

Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.

<a href="https://skillzs.dev/skills/davidlee/doctrine/canon">View canon on skillZs</a>