deepseek-pressure-test
Skill4 starsChanged 20 days agoMITView on GitHub
Adversarial review before you commit. Before a route change, paid CPU or GPU compute, or a worker handoff, an independent worker from another lab tries to kill the route. The output is a falsifiable decision (proceed, measure, park, kill or human review) plus the hidden assumption that could make it wrong. Advisory only.
In this repository34 other files it gives its agents.
- AGENTS.mdAGENTS.md
- accepted-proof-exact-retuneSkill
- beachhead-ratchetSkill
- circuit-edit-integration-safetySkill
- computer-algebra-oraclesSkill
- constraint-construction-searchSkill
- cross-model-review-gateSkill
- ecdsafail-fastevalSkill
- ecdsafail-lowq-playbookSkill
- ecdsafail-oracle-loopSkill
- ecdsafail-pareto-ratchetSkill
- ecdsafail-scratch-dietSkill
- elicit-researchSkill
- exact-search-engineeringSkill
- fast-lean-proof-loopSkill
- hard-problem-contractSkill
- invariant-chartSkill
- jev-second-perspectiveSkill
- lean-arklib-proofsSkill
- lean-finite-certificate-splittingSkill
- mission-commandSkill
- nasqret-playbookSkill
- pip-disciplineSkill
- portable-source-assertionsSkill
- proximity-prize-campaignSkill
- proximity-prize-frontier-archaeologySkill
- research-loopSkill
- route-compare-probe-disciplineSkill
- rust-archive-build-bindingSkill
- send-to-claude-fableSkill
- shot-autopsySkill
- squares-in-squares-campaignSkill
- verify-paired-native-measurementsSkill
- verify-research-transferSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
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 9
Use before route changes, paid compute, or directed worker handoffs.
≈ “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
External pressure-test worker
Use before route changes, paid compute, or directed worker handoffs.
Contract
The worker must try to kill the route first. It is advisory only.
Prompt shape
You are an advisory pressure-test worker for a public benchmark coordination
harness. This is resource-estimation work, not target selection, exploitation,
or key recovery.
Given the route packet below:
1. Try to kill the route.
2. Ask one sharp next question.
3. Propose one creative bounded alternative.
4. Return a falsifiable decision: proceed, measure, park, kill, or request human review.
5. State the hidden assumption that could make you wrong.
Do not request private endpoints, raw logs, keys, unreleased nonces, or submit
authority.Scoring
The research lead scores the response before acting:
- intelligence,
- creativity,
- evidence discipline,
- made-up-risk,
- usefulness,
- whether claims are source-checkable.
Output
- Decision:
- Best objection:
- Next question:
- Bounded alternative:
- Hidden assumption:
- Confidence:
- Required verification:
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.