agentleFS
mahumadadmahumadad/quarry/skills/engage/security-policy-parser/SKILL.md

security-policy-parser

Skill1 starsChanged 2 days agoMITView on GitHub

Parse and extract actionable constraints from a target's bug bounty or vulnerability disclosure policy before planning any testing phase. Identifies: test account policy (provided vs. researcher-owned), explicit exclusions, SLA/payment terms, safe harbor language, and scope boundaries that affect engagement planning. Use after scope is defined and before recon begins. Keywords: bug bounty policy, VDP, vulnerability disclosure, scope rules, test accounts, safe harbor, responsible disclosure, program rules.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is 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.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If test accounts = "researcher-owned" AND no self-service signup exists → flag that authenticated testing requires commercial onboarding or …

    line 71
  2. Very specific:

    If exclusions mention specific techniques (e.g., "no automated scanning") → annotate in scope.json rules.

    line 74
  3. Distinctive:

    If rewards are $0 (VDP only) → adjust ROI expectations.

    line 76

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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.

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

Signals

  • Engagement targeting a bug bounty or VDP program.
  • Need to determine whether authenticated testing is feasible before starting recon (credential availability affects phase planning).
  • Need to confirm safe harbor, exclusions, and reporting requirements.

Prerequisites

  • URL of the target's security/bug-bounty policy (from scope-discipline or the engagement brief).

Methodology

  1. Fetch the policy page:
    curl -s "https://target.example.com/security-policy" > policy-raw.html
    # strip HTML for readable text
    lynx -dump -nolist policy-raw.html > policy-text.txt
    # or
    cat policy-raw.html | sed 's/<[^>]*>//g' | tr -s ' \n' > policy-text.txt
  2. Extract key fields — search for these patterns:

    FieldSearch termsImpact on engagement
    Test accounts"own account", "test account", "we provide", "sandbox"Determines if auth testing needs coordination
    Exclusions"out of scope", "excluded", "do not", "prohibited"Hard limits on techniques
    Safe harbor"legal action", "law enforcement", "good faith"Engagement risk assessment
    Rewards"$", "bounty", "reward", "payment"ROI estimation
    SLA"business days", "response time", "triage"Reporting timeline
    Reporting format"encrypt", "PGP", "format", "steps to reproduce"Report template requirements
    Domains"in scope", "*.domain", "assets"Cross-check against scope.txt
  3. Produce a policy summary — structured output for the engagement state:
    ## Policy Summary: [Target]
    - **URL:** [policy URL]
    - **Last updated:** [date if stated]
    - **Safe harbor:** [yes/no + quote]
    - **Test accounts:** [provided / researcher-owned / unclear]
    - **Account creation path:** [self-service / sales / invitation / unclear]
    - **Exclusions:** [list]
    - **Reward tiers:** [table]
    - **SLA:** [triage days / payment days]
    - **Report requirements:** [encryption, format, etc.]
    - **Domains in scope:** [list, cross-checked with scope.txt]
  4. Flag blockers for phase planning:
    • If test accounts = "researcher-owned" AND no self-service signup exists → flag that authenticated testing requires commercial onboarding or vendor coordination BEFORE planning auth phases.
    • If exclusions mention specific techniques (e.g., "no automated scanning") → annotate in scope.json rules.
    • If rewards are $0 (VDP only) → adjust ROI expectations.
  5. Store in engagement state: Save the summary as report/policy-summary.md and update .pentest-state/scope.json with any rules extracted.

Validation Gate

  • Policy URL fetched and readable text extracted.
  • Test account availability determined (provided / own / unclear).
  • All exclusions captured and reflected in scope rules.
  • Safe harbor language confirmed present (or its absence noted as engagement risk).
  • Account creation path documented — if no self-service, the blocker is flagged before recon completes.

Gotchas

  • Policies change. Always fetch fresh at engagement start — don't rely on cached or remembered versions.
  • "Use your own account" does NOT mean "create an account for free" — if there's no self-service signup, the researcher must subscribe or coordinate with the vendor. This distinction blocked authenticated testing in the Healthatom engagement for 8+ phases.
  • Some policies list assets in a machine-readable format (JSON, YAML in .well-known/security.txt or HackerOne scope API). Prefer structured sources over scraping HTML.
  • "Web/API vulnerabilities" in scope usually covers AI/ML endpoints too, but confirm if there's ambiguity.
  • VDPs (no reward) still provide safe harbor — the legal protection has value even without payment.

Chain Anchors

Policy analysis feeds into:

  1. Engagement planning — credential strategy, phase ordering
  2. Report formatting — match the program's expected format
  3. ROI estimation — reward tiers × finding severity × competition

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.