Security
Model trust boundaries, trace untrusted input to privileged actions, harden controls, and verify fail-closed behavior.From its SKILL.md
npx -y skills add menot-you/n-soul --skill securityAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 18 days oldThe repository was created 18 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 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.
SKILL.md
2.0 KB, 371 tokens by cl100k_base, as published. Nobody here has run it
Security
Use for auth, identity, secrets, untrusted input, network exposure, data mutation, migrations, infrastructure, release, or destructive actions.
Inputs to lock
- Assets and security goals.
- Actors, identities, trust boundaries, and entry points.
- Privileged actions and data stores.
- Current controls, failure history, and accepted risk.
- Authority and rollback for any change.
Method
- Draw the data and authority flow across every trust boundary.
- Enumerate attacker-controlled inputs, credentials, tokens, files, messages, and network paths.
- Trace each input to a privileged sink. Record validation, authorization, isolation, logging, and rate limits along the path.
- Identify abuse cases: privilege escalation, confused deputy, replay, injection, path escape, unsafe deserialization, race, resource exhaustion, and secret disclosure.
- Confirm least privilege, explicit allow rules, secure defaults, bounded resources, and fail-closed behavior when dependencies are missing.
- Bind approval to the exact candidate and prevent substitution between review and the action.
- Keep secrets out of command lines, logs, generated artifacts, and durable handoffs.
- Add negative tests at each boundary and prove denied actions cause no state change.
- Verify audit records are useful without leaking sensitive data.
- Define rollback, recovery, and post-action readback before high-impact action.
Findings
For every finding include asset, entry point, broken control, privileged-action sink, realistic impact, proof, and a clean remediation. Rank severity from demonstrated reach and preconditions, not alarm language.
Do not approve a high-impact action while identity, candidate binding, rollback, or independent verification remains ambiguous. Surface one real blocker and one safe next action; never bury security uncertainty in a long option list.
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.
Gives 0 of the 12 instructions most security skills give in 371 tokens
Counted across 648 of the 828 authors here whose files we hold, read 2026-08-07
- Parameterize all database queriesin 68 of 648, across 51 files
- Hash passwords using bcrypt, scrypt, or argon2in 49 of 648, across 36 files
- Apply rate limiting to authentication endpointsin 48 of 648, across 24 files
- Configure security headersin 35 of 648, across 19 files
- Validate all inputsin 32 of 648, across 24 files
- Validate all external input at the system boundaryin 29 of 648, across 19 files
- Run containers as a non-root userin 28 of 648, across 15 files
- Use httponly secure samesite cookies for sessionsin 26 of 648, across 15 files
- Run dependency audits before every releasein 21 of 648, across 10 files
- Encode output to prevent cross-site scriptingin 21 of 648, across 11 files
- Copy dependencies before source codein 20 of 648, across 9 files
- Store secrets in environment variablesin 20 of 648, across 18 files
Said here and by no other author read
- draw data and authority flow across trust boundaries
- enumerate attacker-controlled inputs and network paths
- trace each input to a privileged sink
- confirm least privilege and fail-closed defaults
- bind approval to the exact candidate
- add negative tests at each boundary
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.