execution-policy-kernel

execution-policy-kernel is a cursor rule for Cursor from hmj1026/dhpk. It costs 3 tokens per session (682 once invoked), scanned A, original, MIT.

A short safety and routing policy for coding-agent work. It tells the agent to stay within authorized files and tasks, protect secrets, preserve existing uncommitted changes, and follow the project's dispatch rules.

In plain words
What is it for?
Use it as an always-visible pre-check before loading more detailed policies or deciding how an agent request should be routed.
Why use it?
It reduces the risk of an agent overwriting a developer's work, exposing sensitive data, or taking an unapproved external action. It also provides a fixed boundary for interpreting commands.

Cursor rule for Cursor

Written for Cursor: a Cursor rule (.mdc).

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add rules/hmj1026/dhpk/execution-policy-kernel
Clone the repo
git clone --depth 1 https://github.com/hmj1026/dhpk

Made for: Cursor.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for execution-policy-kernel

README.md
[![agentmods](https://agentmods.dev/badge/rules/hmj1026/dhpk/execution-policy-kernel.svg)](https://agentmods.dev/rules/hmj1026/dhpk/execution-policy-kernel)
Your own site
<a href="https://agentmods.dev/rules/hmj1026/dhpk/execution-policy-kernel"><img src="https://agentmods.dev/badge/rules/hmj1026/dhpk/execution-policy-kernel.svg" alt="Measured on agentmods" height="20"></a>
Per session 3 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 682 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00003 $0.00682
Opus 5 $0.00002 $0.00341
Sonnet 5 $0.00001 $0.00136
Haiku 4.5 $0.00000 $0.00068

Measured today against content hash 0e96a19cdbc5, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

execution-policy-kernel scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

cursor/rules/execution-policy-kernel.mdc · 69 lines

How it starts

The opening of the file, as written. The whole thing — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Execution Policy Kernel

This is the short, always-visible safety kernel. Read it before loading a conditional stack, version, review, or OpenSpec reference. The full policy in execution-policy.md remains the single source of truth for routing precedence, dispatch selection, and reviewer closure; this file does not duplicate its tables.

Safety and authorization

  • Work only in the user-authorized repository, files, and task scope. A new external side effect, provider, message, or materially different target needs explicit authorization.
  • Preserve existing dirty worktree changes. Do not reset, checkout, stash, delete, overwrite, or auto-commit user work; report ownership ambiguity before editing an overlapping path.
  • Treat secrets, credentials, tokens, private account identifiers, and raw user data as redacted evidence. Do not paste them into logs, prompts, or generated artifacts.

Invocation and route boundary

Command entry points parse flags once through the immutable skills/flow-drive/scripts/route-result.js boundary. Downstream policy consumes that route result and must not reconstruct precedence from the cleaned query. The target invocation class still applies: an explicit-only target is presented with its exact command form rather than called through a delegated Skill handoff.

Completion boundary

Do not claim completion from intent, a successful dispatch call, or a source scan alone. Completion requires the requested scope, actual verification evidence, and an explicit record of skipped, unavailable, blocked, or environment-dependent checks. A pending reviewer, unresolved gate, or unverified runtime premise remains open.

OpenSpec planner gate

Before the first write wave of an existing OpenSpec apply, count the unchecked tasks in the task artifact. If there are >=2 unchecked tasks (two or more), a planner dispatch is mandatory before any writer starts. Record the planner's actionable handoff with the dependency order, the exact owner and write scope for each task, and the next checkpoint; an unrecorded or incomplete planner handoff does not authorize a write wave. If exactly one clear task remains, record planner=skipped and continue through the normal decision, writer, and verification gates. This planner gate is a lifecycle invariant: it remains active when orchestration_dispatch=off, which disables implementation worker/reasoner routing but does not bypass mandatory pre-write planning or verification.

Read the full file on GitHub · 69 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. today First seen · 69 lines · 3 tokens per session scan A 0e96a19cdbc5

Subscribe to this mod's changes

execution-policy-kernel is a cursor rule published in the GitHub repository hmj1026/dhpk (2 stars, last pushed yesterday), licensed MIT. It adds 3 tokens to every session and 682 once invoked, about $0.0000 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-05.