principle-independent-verdict SKILL.md skillA

principle-independent-verdict from vraj-ai/skills is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add vraj-ai/skills/principle-independent-verdict@v2.0.0

Pins v2.0.0 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
vraj-ai/skills
Path
principles/principle-independent-verdict/SKILL.md
Versions recorded
3
Latest version
v2.0.0, committed Oct 27, 2026
First version recorded
Oct 14, 2026
Size
2,301 bytes
Repository stars
3
License
MIT

Only a verdict from a context that did not write the code counts — an author checks what it meant to build, not what it built, and an author-verified claim on its own diff is not enough. Apply before trusting any review, CI-green diff, self-certification, or self-verified work on a nontrivial change, including verdicts carried across rebases where the patch may have drifted.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# Principle: the independent verdict

An author checks what it meant to build. A verifier checks what it built. Those
are different objects, and only the second one protects a merge.

## The rule

A review, a verification pass, or a "tested" claim on a nontrivial change
counts when the examining context did not write the examined code: a separate
agent, a separate session, a CI job, or a human reviewer. Self-verification is
still required — it just isn't a verdict. It is the author's own check before
requesting one.

Green CI and an approving bot review are supporting evidence, not a verdict on
behavior: neither runs the change the way a user would. The strongest
independent verdict runs the real artifact per principle-real-artifact.

## Verdicts are bound to a patch, not a branch

A verdict covers the code the verifier saw. A rebase, retarget, or new push
moves the head to a new SHA without failing any check. Before trusting a
verdict made at an older SHA:

1. Compare the patch-id of the whole diff under review —
   `git diff <base>...<sha> | git patch-id` — at the verdict SHA and the
   current head. The head commit's patch-id misses an earlier commit drifting
…

Read the whole file at its exact version.

How to install

No tool to install first
claude plugin marketplace add https://markdownregistry.com/mp/a/art_jiovzb7cwcs2seaf/5979db2dc2042ac944fc9e9831397cdbbece2b77f48e2aefd8a93fc19a74ced1.json

Then /plugin install principle-independent-verdict@markdownregistry-jiovzb7cwcs2seaf in Claude Code. It installs v2.0.0, the one graded A here, pinned to the commit it was graded at, and leaves it there.

The exact contents, by fingerprint
mdr add vraj-ai/skills/principle-independent-verdict@sha256:5979db2dc2042ac9

The line at the top of this page pins the label v2.0.0, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_jiovzb7cwcs2seaf.svg)](https://markdownregistry.com/a/art_jiovzb7cwcs2seaf)

3 badge image fetches in 30 days. What the badge shows.

Ask from your agent
claude mcp add --transport http markdownregistry https://markdownregistry.com/mcp/registry

Then ask it: what changed in vraj-ai/skills/principle-independent-verdict since the version I have? It reads every saved version here, no key and no account. Any agent that speaks MCP can add it.

Got a copy of principle-independent-verdict from somewhere else? Ask it: where did my principle-independent-verdict come from, and what was changed? It names the public original, the version yours is closest to and every changed line. The text is not kept.

Versions

The registry has recorded 3 versions of this file, between Oct 14, 2026 and Oct 27, 2026, about one every 6 days.

Watch this file (sign in, one email per new version)

versioncommittedcommitsizeaudit
v2.0.0 latest2026-10-27 e242bc0 2,301 BA view · diff
v1.0.02026-10-14 6cb60ee 2,186 BA view · diff
v1.0.02026-10-14 6bd9b35 2,009 BA view
Keep it current by itself, install once

Committed this skill to a repository? Install the markdownregistry skill updates app on it, or on all of them. When the registry saves a newer version, a pull request brings your copy to it within minutes, the newest version first, then what changed and any check it newly trips. A copy you edited is never touched, and nothing is deleted.

Install on GitHub

It asks to read and write contents and pull requests, and to read metadata. No workflow file, no Actions setting, nothing runs on your runners. Uninstall it and everything the registry kept about your repository is deleted.

Keep it current by itself

Committed this skill to a repository? Add one workflow file, and when a newer version is saved, a pull request brings your copy to it, the newest version first, then what changed and any check it newly trips. A copy you edited is never touched, and nothing is deleted.

mkdir -p .github/workflows && curl -fsSL https://markdownregistry.com/ci/skill-updates.yml -o .github/workflows/skill-updates.yml

It runs daily on GitHub Actions, no key and no account. Read the workflow before you add it.

Check a whole project

From the root of your project, one line hashes each installed skill's SKILL.md, your AGENTS.md and your CLAUDE.md, and asks the registry once. Only the fingerprints and paths are sent.

sha256sum .claude/skills/*/SKILL.md AGENTS.md CLAUDE.md | curl -s --data-binary @- https://markdownregistry.com/api/v1/check-copies

In CI, add --fail-with-body: the answer is 409 when any file is behind, so the step fails and still prints which files.

sha256sum .claude/skills/*/SKILL.md AGENTS.md CLAUDE.md | curl -s --fail-with-body --data-binary @- https://markdownregistry.com/api/v1/check-copies

No sha256sum on your machine? shasum -a 256 prints the same lines.

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (2301 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

vraj-ai/skills · 3 stars · license MIT · pushed 2026-11-11 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_jiovzb7cwcs2seaf
GET https://markdownregistry.com/api/v1/resolve?ref=vraj-ai/skills/principle-independent-verdict
GET https://markdownregistry.com/api/v1/blob/5979db2dc2042ac944fc9e9831397cdbbece2b77f48e2aefd8a93fc19a74ced1

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from vraj-ai/skills

verify-skills SKILL.md skill
vraj-ai/skills · .cursor/skills/verify-skills/SKILL.md · Prove the vskills repository by linting every skill's frontmatter and relative cross-references, then running the test…
v1.1.0 · audit A · 3 stars
AGENTS.md AGENTS.md file
vraj-ai/skills · AGENTS.md
git:20261025.fd2cd15 · audit A · 3 stars
land SKILL.md skill
vraj-ai/skills · delivery/land/SKILL.md · Drive open PRs to merged. Re-establish each PR's real state, resolve review threads, conflicts, failing checks, and CI…
v1.7.0 · audit A · 3 stars
ship-record SKILL.md skill
vraj-ai/skills · delivery/ship-record/SKILL.md · Check, attest, and export the ship record every delivery skill shares. Reads the record at $(git rev-parse…
v1.3.0 · audit A · 3 stars
ship SKILL.md skill
vraj-ai/skills · delivery/ship/SKILL.md · Find, pick up, implement, or execute a feature, bug, task, ticket, or ready issue end to end and open a verified…
v4.8.0 · audit A · 3 stars
ai-subscription-unit-economics SKILL.md skill
vraj-ai/skills · pipeline/ai-subscription-unit-economics/SKILL.md · Pricing, limits, and margin gates for a product whose cost of goods is model inference — compute cost per action, cost…
v1.0.0 · audit A · 3 stars
audit SKILL.md skill
vraj-ai/skills · pipeline/audit/SKILL.md · Audit a whole codebase rather than a single diff — establish the boundary and baseline, sweep layer by layer (entry…
v1.0.0 · audit A · 3 stars
delivery-constraints SKILL.md skill
vraj-ai/skills · pipeline/delivery-constraints/SKILL.md · Deliver a ticket whose delivery path is decided by something other than the code — a token/spend cap, a live data…
v1.1.0 · audit A · 3 stars
implementation-tdd SKILL.md skill
vraj-ai/skills · pipeline/implementation-tdd/SKILL.md · Strict one-ticket TDD for implementation work — write the failing test first at the real production seam, lock the…
v1.1.0 · audit A · 3 stars
subagent-delegation SKILL.md skill
vraj-ai/skills · pipeline/subagent-delegation/SKILL.md · Fan work out to subagents safely — delegate independent packages to workers without losing ownership: draw provably…
v1.1.0 · audit A · 3 stars
principle-bounded-repair SKILL.md skill
vraj-ai/skills · principles/principle-bounded-repair/SKILL.md · Bound every repair loop — retrying an identical failure the same way burns budget without generating information. Set a…
v1.0.0 · audit A · 3 stars
principle-evidence-over-report SKILL.md skill
vraj-ai/skills · principles/principle-evidence-over-report/SKILL.md · When a report and the repository disagree — a worker claims success with no commit, a tracker says done but the gate is…
v1.0.0 · audit A · 3 stars

Every file in vraj-ai/skills

Browse by kind, by grade A, or by owner.