PromptInjection
Skill2 starsChanged 44 days agoMITView on GitHub
Assess authorized AI features as data-flow and tool-permission boundaries. Use controlled canaries and test accounts; do not request real secrets or unauthorized tool actions. Do not promote an answer-style change to a confirmed vulnerability unless it causes a reproducible, unauthorized effect.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- mj-deving/pai-skillstheir own version, first seen Apr 27, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- mj-deving/pai-skills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository8 other files it gives its agents.
- bug-bountySkill
- APIAssessmentSkill
- bb-methodologySkill
- BusinessLogicSkill
- EvidenceValidationSkill
- HuntPlanningSkill
- ReconSkill
- WebAssessmentSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 6
Place a harmless canary instruction in one lower-trust surface and compare against a baseline and an unrelated control.
- Distinctive: line 8
Verify any tool effect from independent logs or state reads.
- Distinctive: line 12
Do not promote an answer-style change to a confirmed vulnerability unless it causes a reproducible, unauthorized effect.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
AI and Prompt Injection Assessment
Assess authorized AI features as data-flow and tool-permission boundaries. Use controlled canaries and test accounts; do not request real secrets or unauthorized tool actions.
- Map the model's trusted instructions, user input, retrieval sources, browser content and available tools.
- Place a harmless canary instruction in one lower-trust surface and compare against a baseline and an unrelated control.
- Measure whether the model merely quotes the canary, follows it in its answer, or crosses a tool-permission boundary. These are different outcomes.
- Verify any tool effect from independent logs or state reads. A model's claim that it acted is not proof.
- Repeat with distinct benign canaries to test robustness while keeping attempts bounded by the operator's budget.
- Record the exact trust boundary, input source, model configuration, observed response and required human review.
Do not promote an answer-style change to a confirmed vulnerability unless it causes a reproducible, unauthorized effect.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.