SkillAgentSearch skills...

first-principles-review

Use when asked for first-principles or Occam's-razor review, or when high-risk decisions involve competing constraints, fallback growth, duplicate owners, or architecture direction risk. Ordinary bug fixes stay on the fast path.

Install / Use

npx skills add GanyuanRan/Aegis --skill first-principles-review

Installs into whichever agent you are using.

About this skill
📄

SKILL.md

Installable skill definition

Quality Score

92/100

Category

Automation

Supported Platforms

Universal

Our assessment of first-principles-review

first-principles-review scores 92/100 on our quality scale, 895th of 2,780 Automation skills we index (top 33%).

Its SKILL.md is 9.7 KB long, well organised into 9 sections with 5 code examples: a thorough specification that gives an agent plenty to work with.

With 1,273 GitHub stars, it is one of the more widely adopted skills in the catalogue.

Substance
29/30
Structure
20/20
Description
15/15
Adoption
13/20
Freshness
15/15

Maintenance, license and trust

  • The repository was last updated about 2 months ago, so first-principles-review is actively maintained.
  • Our last check on 2026-10-27 found the source still online.
  • It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
  • Its trust signals score 100/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.

Safety scan

No issues found

Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.

AI review by kimi-k2.7-code on 2026-10-04. Automated pattern scan on 2026-10-04. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.

first-principles-review compared with similar skills

All 4 of these similar skills score higher than first-principles-review; compare them before choosing.

SkillScoreStarsUpdatedFormat
first-principles-review (this skill)by GanyuanRan921.3k53d agoSKILL.md
Agent-Reachby Panniantong100112.3k8d agoCLAUDE.md
rufloby ruvnet10081.9ktodayCLAUDE.md
Claude-Code-Game-Studiosby Donchitos10039.7k7d agoCLAUDE.md
Scraplingby D4Vinci100103.6k2d agoMCP Server

Frequently asked questions

How do I install first-principles-review?
Run npx skills add GanyuanRan/Aegis --skill first-principles-review. The install tabs above show the steps for each supported agent.
Which AI agents does first-principles-review work with?
It is written for Universal, as a SKILL.md file. Other agents that read the same format can often use it too.
Is first-principles-review safe to use?
Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It is MIT-licensed and scores 100/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is first-principles-review still maintained?
The repository was last updated about 2 months ago, so first-principles-review is actively maintained.

name: first-principles-review description: "Use when asked for first-principles or Occam's-razor review, or when high-risk decisions involve competing constraints, fallback growth, duplicate owners, or architecture direction risk. Ordinary bug fixes stay on the fast path."

First Principles Review

Purpose

Use this as a lightweight decision review before another Aegis workflow makes a directional choice. It is a compositional skill, not a standalone workflow.

Do not replace brainstorming, systematic-debugging, writing-plans, requesting-code-review, or verification-before-completion. Use it to clean the decision surface those skills will act on.

When this review materially changes the direction, surface Aegis Visibility in natural prose: name the first principle, dropped assumption, smallest sufficient path, or owner / retirement falsifier that changed the decision. Keep it advisory and task-specific; do not turn the lens into approval authority or a generic skill trace.

Use When

  • The user asks for first principles, first-principles thinking, or Occam's razor.
  • A design, plan, or fix has multiple plausible paths and unclear selection criteria.
  • The task has ambiguous goals, competing constraints, or product/architecture direction risk.
  • Debugging is drifting into repeated fixes, fallback growth, duplicate owners, consumer-side patches, or "just add another branch" reasoning.
  • A review finds that the implementation may be locally correct but directionally wrong.

Do Not Use

  • Simple Q&A, status checks, tiny wording/config edits, or clearly bounded single-owner changes.
  • Mechanical execution of an approved plan unless a new directional conflict appears.
  • As a required step for every task, every turn, or every TDD cycle.

Five-Line Review

Answer only what is needed, usually in five short lines:

First Principle: What irreducible outcome must this satisfy?
Non-negotiables: What constraints cannot be broken?
Assumptions to Drop: What is habit, inherited shape, or unproven preference?
Smallest Sufficient Path: What is the least complex path that satisfies the first principle?
Escalation Signal: What finding would require spec/design/architecture review?

When the direction depends on a new mechanism or an unfamiliar domain, insert one optional line between path and escalation:

Known Prior Art: proven external pattern worth adopting or adapting to project
constraints (cite source), or `unknown` when precedent cannot be verified here

For repair choices, "smallest" means smallest sufficient stable repair, not the smallest textual diff:

Minimality Check:
- Smallest textual diff:
- Correct owner:
- Bug class fixed:
- New branch/fallback added:
- Old path retired or scheduled:
- Verdict: sufficient repair | local patch | needs first-principles review

Decision Hygiene Review

Use this escalation only when a design, fix, or plan needs endorsement before it is written into a spec or implementation plan.

Escalate from the five-line review when any of these risk signals appear:

  • multiple plausible paths and no clear selection criteria
  • a new owner, duplicate owner, fallback, adapter, or compat-only carrier
  • an old path that may need delete-first handling or a retirement trigger
  • an unverified assumption that the proposal depends on
  • user language such as "more elegant", "long-term stable", "first principles", or "Occam"
  • a plan could encode the wrong owner, abstraction, compatibility boundary, or retirement schedule
  • an existing object, behavior, responsibility, contract, or relationship may be reinterpreted, narrowed, replaced, or retired, and the proposal could lose a legitimate role or explicit reference while removing invalid authority

Use this compact shape:

First-principles invariants:
- Non-negotiable goal:
- Non-negotiable constraints:
- Historical assumptions to delete:

Bounded preservation reminder:
- Evidence-backed behavior that must remain correct:
- Highest-risk counterexample:
- Material unknown / uninspected surface:
- Known explicit anchors / upstream-downstream refs and disposition:
- Role-before-value ambiguity, if any:

Owner / retirement matrix:
- New canonical owner:
- Old owner:
- Compat-only carrier:
- Delete-first / retirement trigger:

Falsification matrix (evaluate in order; stop at the first failing gate):
- Gate 0 - Premise evidence (stage-graded): design stage accepts spec/logic
  refs; implementation/runtime stage requires log, telemetry, or test
  evidence. No evidence -> park as watch-listed, do not enter value
  evaluation.
- Gate 1 - Decidability: no obtainable evidence to judge it -> park as a
  watched falsifier, do not act.
- Gate 2 - Value (any one): lowers future fix probability / shrinks the
  solution set / exposes a missing actionable acceptance criterion / the same
  broken assumption is reused elsewhere. Passing here still requires
  gates 3-5.
- Gate 3 - Cost match: change cost vs business priority, costed at the
  current stage; a valid but deferred counterexample is recorded in the owner
  doc, never parked orally.
- Gate 4 - Goal regression: the constrained solution still meets the original
  goal; re-read the accepted-constraint list at existing checkpoints (design
  review, plan approval, pre-completion verification).
- Gate 5 - Minimize then classify: boundary-shaped destructive claims must be
  minimized first; premise attacks classify as destructive directly.
  Classes: destructive (refactor/retirement track) | supplementary (boundary
  constraint) | watch-listed (falsifier). Conflicting counterexamples are
  decided by business-goal ranking, never by stacking constraints.
- Gate 6 - Adoption trace: the cheapest checkable form (regression test for
  destructive; contract, assertion, or checklist line for supplementary),
  written into the existing owner doc.

Verdict:
- Adopt / revise / reject / needs evidence:
- Blocking gaps:
- Next evidence:

Architecture Integrity Lens

Use this narrower lens when a proposal is executable but may still encode the wrong owner, abstraction, contract boundary, or retirement path. It is advisory method-pack output and may be embedded inside Decision Hygiene Review when that is enough.

Trigger it when any of these appear before approach selection, task decomposition, review, or completion-risk reporting:

  • responsibilities may overlap or a canonical owner is unclear
  • the smallest diff adds a caller-side fallback, guard, adapter, or compat-only carrier
  • an existing source-of-truth or contract could solve the class of problem at a higher level
  • a stale owner, fallback, or old path may keep carrying real logic
  • the work makes a long-term stability, "cleaner architecture", or higher-level simplification claim

Use this compact shape:

Architecture Integrity Lens:
- Invariant: What must remain true for the system to be coherent?
- Canonical owner / contract: Which owner, contract, or source-of-truth should carry the behavior?
- Responsibility overlap: What duplicate owner, caller-side patch, fallback, or stale path might still carry real logic?
- Higher-level simplification: Can the problem be solved at the owner / contract / source-of-truth layer instead of by another local branch?
- Retirement / falsifier: What old path retires, or what evidence would disprove this architecture judgment?
- Responsibility / capability boundary: Which invalid authority retires, and
  does the same carrier still serve a separately evidenced legitimate role?
- Verdict: proceed | revise design | split owner | return to baseline | needs ADR/baseline sync

Bounded preservation reminder is a risk-triggered reasoning aid, not a universal artifact or an exhaustive behavior inventory. Inspect the smallest relevant contract, consumer, test, or history evidence. Preserve, rebind, retire, or reject each known explicit reference; state unresolved relationships as unknown instead of re-inferring them or claiming semantic completeness. The Method Pack does not build an authoritative relationship graph, prove referential integrity or input lineage, calculate complete behavior coverage, or issue a runtime gate.

Do not run this lens for every low-risk task. If it does not change the decision surface, return to the active workflow immediately.

Composition

  • With brainstorming: run before approach selection when the request is broad, ambiguous, likely to inherit a poor product shape, or involves owner / retirement / fallback / adapter risk. Use Decision Hygiene Review or the narrower Architecture Integrity Lens before recommending or selecting an approach when those signals appear.
  • With systematic-debugging: run after evidence shows repeated fixes, fallback growth, duplicate owners, or consumer-side patching.
  • With writing-plans: run before task decomposition when the plan could encode the wrong owner, abstraction, compatibility boundary, fallback, adapter, or retirement schedule. If the approved spec did not already cover this, use Decision Hygiene Review or the Architecture Integrity Lens before writing tasks.
  • With requesting-code-review: run when review should check direction and owner integrity, not just code quality.
  • With verification-before-completion: use only to name residual directional risk. It does not grant completion authority.

Boundaries

  • Prefer evidence from current project files, baseline docs, tests, logs, and user requirements. If evidence is missing, mark the line as unknown rather than inventing a principle.
  • Keep the result advisory. This skill may recommend escalation, but it does not create authoritative GateDecision, PolicySnapshot, or completion authority.
  • If the five-line review does not change the decision surface, return to the active workflow immediately.

Related Skills

View on GitHub
GitHub Stars1.3k
CategoryAutomation
Updated1mo ago
Forks64

Languages

Python

Trust signals

100/100

From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.

No cautions
first-principles-review — Universal Skill: Install & Safety Check | SkillAgent