agentleFS
jaskaranhundaljaskaranhundal/usap-skills/webapp-security/CLAUDE.md

usap-skills / webapp-security

CLAUDE.md4 starsChanged 2 months agoApache-2.0View on GitHub

The webapp-security/ domain holds skills that triage, classify, and assess web-application and API security findings. Scope is the runtime surface — production webapps and API endpoints — and the OWASP Top 10 2025 + OWASP API Security Top 10 are the primary taxonomies. Build-time AppSec (SAST, dependency analysis, supply-chain) lives in appsec-devsecops/ and is not duplicated here. Skills in this domain produce structured triage payloads, not remediation actions. Mutating recommendations (WAF rule changes, account disablement, schema-rewrite proposals) are surfaced via…

The file

We split this CLAUDE.md into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    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
  1. Distinctive:

    All paths are relative from the repository root.

    line 24
  2. Distinctive:

    All three are stdlib-only and emit USAP's 11-field output contract.

    line 36

≈ “Common” is an estimate: at least 1 in 5 of the 27,641 CLAUDE.md 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
CLAUDE.md

Webapp Security Domain — CLAUDE.md

Purpose

The webapp-security/ domain holds skills that triage, classify, and assess web-application and API security findings. Scope is the runtime surface — production webapps and API endpoints — and the OWASP Top 10 2025 + OWASP API Security Top 10 are the primary taxonomies. Build-time AppSec (SAST, dependency analysis, supply-chain) lives in appsec-devsecops/ and is not duplicated here.

Skills in this domain produce structured triage payloads, not remediation actions. Mutating recommendations (WAF rule changes, account disablement, schema-rewrite proposals) are surfaced via human_approval_required: true and routed to cs-appsec-engineer (the orchestrator) or cs-incident-responder (active-exploit case).

Subdomains covered:

  • Webapp finding triage (which OWASP category, what severity, who consumes it next)
  • OWASP Top 10 2025 classification with confidence scoring
  • API security posture assessment (OWASP API Top 10, BOLA / mass assignment / broken auth)

Skills Catalog

SkillSlugPrimary ToolFrameworks (frontmatter)
webapp-risk-triagewebapp-security/webapp-risk-triagewebapp-risk-triage_tool.pyOWASP A01, A03, A05, A07; MITRE T1190
owasp-top10-classifierwebapp-security/owasp-top10-classifierowasp-top10-classifier_tool.pyOWASP A01–A10 (full set)
api-security-posturewebapp-security/api-security-postureapi-security-posture_tool.pyOWASP A01, A03, A07; MITRE T1078, T1190

All paths are relative from the repository root.


Python Tools Reference

ToolPathPurposeKey Args
webapp-risk-triage_tool.pywebapp-security/webapp-risk-triage/scripts/webapp-risk-triage_tool.pyMap a webapp finding to OWASP + severity, recommend the next USAP skill--input <finding.json>, --output {json,human}
owasp-top10-classifier_tool.pywebapp-security/owasp-top10-classifier/scripts/owasp-top10-classifier_tool.pyClassify a finding description into OWASP Top 10 2025 categories with confidence--input <finding.json>, --output {json,human}
api-security-posture_tool.pywebapp-security/api-security-posture/scripts/api-security-posture_tool.pyScore an API surface (auth, rate limits, identification, mass-assignment) against OWASP API Top 10--input <api-spec.json>, --output {json,human}

All three are stdlib-only and emit USAP's 11-field output contract.


Domain Best Practices

  1. Triage before scanning. A webapp finding without context (target URL, environment, auth state) cannot be prioritised. Always run webapp-risk-triage before invoking a deeper scanner or pentest skill. The triage output names the right downstream skill.
  2. OWASP category is a routing decision, not a verdict. owasp-top10-classifier returns ranked category candidates with confidence. Treat the top match as the routing key, not as a finding statement. The eventual finding statement still requires evidence corroboration.
  3. API auth failures dominate the threat model. OWASP API Top 10's "Broken Object Level Authorization" (BOLA) is the most common real-world issue. api-security-posture scores BOLA visibility first and weights it heaviest in the posture score.
  4. Always emit a next_agents recommendation. Triage skills must point to the next concrete USAP skill — appsec-devsecops/sast-dast-coordinator for build-time, response/incident-classification for production exploit, risk-compliance/risk-threat-modeling for design-stage. Empty next_agents is an anti-pattern in this domain.
  5. human_approval_required: true on any control change. WAF rule injection, schema rewrites, and account state changes are mutating actions — never auto-trigger. Surface them with confidence, evidence references, and a clear escalation owner role.

Workflow Patterns

Workflow 1: Production webapp alert → routing

webapp-risk-triage           (classify finding, score severity, scope blast radius)
       |
       v
owasp-top10-classifier       (refine OWASP category with confidence)
       |
       +--> [A01/A07 + production exploit]  --> response/incident-classification
       |
       +--> [A03/A08 + design issue]        --> risk-compliance/risk-threat-modeling
       |
       +--> [build-time gap]                --> appsec-devsecops/sast-dast-coordinator

Workflow 2: API security review

api-security-posture         (score API surface against OWASP API Top 10)
       |
       v
[posture below threshold]
       |
       +--> appsec-devsecops/secure-sdlc        (design changes)
       +--> appsec-devsecops/security-requirements-review  (PRD updates)
       +--> identity-access/identity-access-risk  (auth model review)

Related Domains

appsec-devsecops/

Build-time AppSec is upstream of this domain. The SAST/DAST coordinator, supply-chain risk, and secure-SDLC skills run in CI; this domain handles their runtime equivalents. Findings flow both ways: a runtime triage may demand a build-time gate, and a CI finding may require runtime monitoring.

response/

Production exploit cases escalate from webapp-risk-triage straight into response/incident-classification. The triage output is structured exactly so the incident-classification skill consumes it without translation.

risk-compliance/

Design-stage webapp issues (architecture, threat-modelling gaps) route to risk-compliance/risk-threat-modeling rather than to incident response. The triage skill chooses which based on the finding's evidence references.


Path Reference

webapp-security/<slug>/
  SKILL.md
  README.md
  references/workflow.md
  assets/templates/output-template.json
  expected_outputs/sample_output.json
  scripts/<slug>_tool.py

Invoke any tool from the repo root:

python3 webapp-security/webapp-risk-triage/scripts/webapp-risk-triage_tool.py --output json
python3 webapp-security/owasp-top10-classifier/scripts/owasp-top10-classifier_tool.py --output json
python3 webapp-security/api-security-posture/scripts/api-security-posture_tool.py --output json

Each tool emits a USAP 11-field contract payload by default. Pass --input <finding.json> to drive the analysis from real data.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.