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.
npx agentmods add rules/prisma/prisma-next/no-pull-request-targetgit clone --depth 1 https://github.com/prisma/prisma-nextWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00027 | $0.00496 |
| Opus 5 | $0.00014 | $0.00248 |
| Sonnet 5 | $0.00005 | $0.00099 |
| Haiku 4.5 | $0.00003 | $0.00050 |
Grade A, and why
no-pull-request-target 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 yesterday.
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.
This is a copy
100% identical to no-pull-request-target — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
What it actually says
Forbidden: pull_request_target
Do not introduce the pull_request_target event trigger in any workflow under .github/workflows/. CI enforces this via scripts/lint-workflow-triggers.mjs, which fails the Lint job if the token appears anywhere in a workflow file outside a YAML comment.
Why
pull_request_target runs in the base repository's trust context — it has secrets and a writable GITHUB_TOKEN — but it fires for PRs from forks. If such a workflow ever executes the PR's code, directly or transitively (e.g. via a shared cache restored from a fork-PR run, or a third-party action that interpolates PR metadata into a shell command), fork-controlled code runs with the base repo's permissions. This is GitHub's "Pwn Request" pattern.
It was the entry point of the May 2026 TanStack supply-chain compromise that published 84 malicious npm versions across 42 @tanstack/* packages.
- TanStack postmortem: https://tanstack.com/blog/npm-supply-chain-compromise-postmortem
- GitHub Security Lab on Pwn Requests: https://securitylab.github.com/resources/github-actions-preventing-pwn-requests/
What to use instead
pull_request— runs with a read-onlyGITHUB_TOKENand zero secrets when the PR comes from a fork (GitHub default; we rely on it).- For commenting on PRs from a privileged side, prefer the
workflow_runtrigger paired with strict input validation, or schedule a job that reads PR metadata via the API rather than running the PR's code.
If you genuinely need it
Don't paper over the lint with a per-file override. Open a PR that:
- Justifies the use case to code owners.
- Demonstrates that the workflow does not check out, build, install, or otherwise execute fork-controlled content.
- Updates
scripts/lint-workflow-triggers.mjsto reflect the new policy (e.g. an explicit allow-list keyed on filename and event declaration shape).
See docs/oss/supply-chain.md for the broader fork-PR posture.
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.
- yesterday First seen · 33 lines · 27 tokens per session scan A 85877984cf07
no-pull-request-target is a cursor rule published in the GitHub repository prisma/prisma-next (421 stars, last pushed 7d ago), licensed Apache-2.0. It adds 27 tokens to every session and 496 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to no-pull-request-target, differing in 0 lines, and is treated as a copy.
Other cursor rules, from other repositories
cli-error-handling
CLI command error handling patterns.
prefer-direct-imports-over-module-mocks
Prefer extracting a testable core over vi.mock / vi.resetModules when unit tests need to reach production logic entangled with config, env, or singletons.
control-plane-descriptors
Control plane descriptor and instance implementation patterns.
directory-layout
Directory layout rules for SQL family vs concrete targets.
family-instance-domain-actions
Family instance domain action implementation patterns.
import-validation
Import validation and package layering rules.