agentleFS
NagiliantNagiliant/Genesis-Legacy-V2/skills/genesis-a4/SKILL.md

A4 — North Star Direction

Skill2 starsChanged 5 months agoApache-2.0View on GitHub

Vision-gap analysis. Compared to the perfect-world North Star, what is missing, drifting, or underbuilt — and what to add or change now to close the gap. Not a passive alignment scorecard.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Write the report to audits/{scope}-a4.md: per-dimension Aligned/Drifting/Missing with gap analysis, compass-question results, actionable …

    line 25

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

A4 — North Star Direction (Gap Analysis)

The vision-gap check — and emphatically not a passive alignment scorecard. A4 asks: compared to the North Star (the perfect-world vision for this feature), what is missing, drifting, or underbuilt — and what should we add, build, or change now to close that gap? For every quality dimension it asks "what would perfect look like here, and what is the delta?" and produces concrete improvements to make, not checkmarks. This lens requires top-tier judgment.

Inputs

  • northstar.md (required — refuse if missing) · the implementation code in scope · the micro specs · design.md
  • The constitution · prior A1–A3 reports · the project mode
  • Accumulated domain-learnings.md from prior A4 runs (extra, project-earned quality expectations)

Process

  1. Read the North Star deeply — internalize every dimension, compass question, and emotional quality; understand what "perfect" means for this feature.
  2. Walk each quality dimension (weighted Critical / Important / Nice). For each: what does perfect look like for this implementation? what is the current state? what is the gap? Score Aligned / Drifting / Missing. For anything drifting or missing, state the specific change that closes the gap. Critical-dimension gaps are urgent.
  3. Walk the compass questions — apply each to the implementation's decisions; flag any that fail, and state what passing would look like.
  4. Assign severity by dimension weight: Critical → High/Critical · Important → Medium · Nice → Low (still fixed).
  5. Make every finding actionable — never "this doesn't feel right," always a concrete change (e.g. "the settings panel shows 12 options at once; the North Star calls for cognitive-load minimization — group into 3 categories of 4").
  6. Capture domain learnings — record domain-specific quality observations that emerged from the work itself, appended to domain-learnings.md, tagged with the feature/domain. They accumulate: each future A4 run reads them as additional expectations. This is how the audit gets smarter about a domain over time.
  7. Write the report to audits/{scope}-a4.md: per-dimension Aligned/Drifting/Missing with gap analysis, compass-question results, actionable findings, captured observations, and an overall direction assessment.

Findings beyond this scope

A4 has a fourth disposition: defer what's blocked — with a concrete pickup path and a named blocker. Any finding suggesting a change beyond this spec's scope is written into the feature's design-inputs.md (a plain markdown journal of spec-amendment inputs). This is not a deferral-with-no-record — it converts the gap into a tracked work item. The open-finding count still reaches zero.

Shared machinery

Finding format, the zero-tolerance loop (even Nice-dimension drift is fixed), and conviction: see docs/the-audit-system.md.

Natural language triggers

- "a4 audit group 1"
- "north star gap analysis"
- "what's missing vs the vision"

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.