Agent action gate

Skill SeezaZzz/agent-action-gate-kit/skills/agent-action-gate

PASS/STOP pre-action receipts for Codex and AI coding agents before edits, handoffs, commits, or publish steps.

Install

npx -y skills add SeezaZzz/agent-action-gate-kit --skill agent-action-gate

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Run a PASS/STOP receipt before risky AI-agent actions such as file edits, tool actions, handoffs, commits, releases, or status answers. Use when the agent must prove sources checked, substrate readiness, deferred scope, and self-challenge before acting.

SKILL.md

2.1 KB, 446 tokens by cl100k_base, as published. Nobody here has run it

Agent Action Gate

Use this skill to convert "think before acting" into a command-backed receipt.

Required Behavior

Before a non-trivial action, run the gate script for the active workspace. Prefer the installed global path:

powershell -NoProfile -ExecutionPolicy Bypass -File "$env:USERPROFILE\.codex\agent-action-gate\scripts\Test-AgentActionGate.ps1" `
  -WorkspaceRoot "<project-root>" `
  -ActionType "<pre-answer|pre-status-report|pre-recommendation|pre-plan|pre-file-edit|pre-tool-action|pre-review|pre-handoff|pre-dispatch|pre-commit|pre-publish>" `
  -RequestedAction "<plain action being taken>" `
  -SkillsUsed "<skills/tools/runbooks used>" `
  -RequiredSubstrate "<sources/state/artifacts consumed>" `
  -SubstrateStatus "<IMPLEMENTABLE|PARTIAL|MISSING|NOT_APPLICABLE>" `
  -CheapestUsefulVersion "<smallest useful valid action>" `
  -DeferredScope "<intentionally deferred work>" `
  -SelfChallenge "<why this is not skipping evidence or accepting a bad option set>" `
  -OwnerFacingOutput "<exact output being prepared>"

If the project has not been initialized, initialize it first:

powershell -NoProfile -ExecutionPolicy Bypass -File "$env:USERPROFILE\.codex\agent-action-gate\scripts\Initialize-AgentActionGate.ps1" -WorkspaceRoot "<project-root>"

Stop Rules

PASS means the pre-action workflow evidence exists.
STOP_* means stop and fix the missing source/skill/substrate/challenge/output before acting.
PARTIAL or MISSING substrate does not authorize hard actions such as file edit, tool mutation, handoff, dispatch, commit, or publish.
The receipt does not grant Git, live-service, destructive, payment, promotion, publish, or owner-only permission.

Reference

Read references/action-gate.md when configuring a new project, explaining the receipt fields, or adapting the gate to a custom workflow.

What ships with it: 2 files

1.8 KB alongside SKILL.md

agents/

openai.yaml266 B

references/

action-gate.md1.6 KB

Gives 0 of the 12 instructions most context ai engineering skills give in 446 tokens

Counted across 1,193 of the 1,976 authors here whose files we hold, read 2026-08-07

Dispatch a fresh implementer subagent per taskin 48 of 1193, across 19 files
Dispatch a final code reviewer after all tasksin 33 of 1193, across 8 files
Provide full task text to the subagentin 30 of 1193, across 9 files
Review spec compliance before code qualityin 27 of 1193, across 10 files
Make the hook script executablein 26 of 1193, across 8 files
Re-snapshot after navigation or DOM changesin 25 of 1193, across 19 files
Read files before editing themin 22 of 1193, across 11 files
Answer subagent questions before proceedingin 22 of 1193, across 7 files
Mark task complete in TodoWrite after approvalin 22 of 1193, across 6 files
Merge hook into existing settingsin 21 of 1193, across 3 files
Ask if installation is global or projectin 20 of 1193, across 2 files
Copy the hook script to target locationin 20 of 1193, across 2 files

Said here and by no other author read

run the gate script before non-trivial actions
stop and fix missing evidence on STOP_* receipts
read the reference when configuring a new project
provide plain action being taken
provide skills or tools used
provide sources or artifacts consumed

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.