/org-memory-compile
[experimental] Compile human-reviewed scan history into local organization memory
$ npx -y skills add allsmog/vuln-scout --agent claude-codeHow it fires
How this command gets triggered: by you, by Claude, or both.
- Fires itselfClaude auto-loads it when your prompt matches the work.
- You can call itInvoke it directly when you want it.
- Slash command
/org-memory-compile
Context preview
What this command does when you run it.
[experimental] Compile human-reviewed scan history into local organization memory
Command definition
org-memory-compile.mdname: org-memory-compile
description: "[experimental] Compile human-reviewed scan history into local organization memory"
argument-hint: "[--privacy open|hashed|strict] [--dry-run] [--allow-commit] [--force]"
allowed-tools:
- Bash
- Read
- Write
- Glob
- Grep
Compile Org Memory
Compile repeated human-reviewed verdicts into `.vuln-scout/org-memory/` so future scans can reuse organization-specific suppressions, confirmed patterns, and reviewer heuristics.
Usage
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/org_memory_compiler.py" \
--project-root . \
--privacy hashedpython3 "${CLAUDE_PLUGIN_ROOT}/scripts/org_memory_compiler.py" \
--project-root . \
--privacy strict \
--dry-runInputs
- `.claude/scan-history/*.json`
- `.claude/rule-stats.json`
- `.claude/review-ledger.json`
- Human-reviewed findings where `verdict` is `verified` or `false_positive`
Outputs
- `.vuln-scout/org-memory/accepted-suppressions.yaml`
- `.vuln-scout/org-memory/confirmed-findings.yaml`
- `.vuln-scout/org-memory/custom-rules/semgrep/*.yaml`
- `.vuln-scout/org-memory/review-patterns.yaml`
- `.vuln-scout/org-memory/manifest.json`
Flags
| Flag | Effect | |------|--------| | `--privacy open` | Store paths and excerpts verbatim | | `--privacy hashed` | Store SHA-256 hashes for paths and excerpts | | `--privacy strict` | Store only rule IDs, verdict counts, and CWE metadata | | `--dry-run` | Print proposed memory without writing files | | `--force` | Allow overwriting strict memory with open privacy | | `--allow-commit` | Do not add `.vuln-scout/org-memory/` to `.gitignore` |
Review Policy
Only human-reviewed provenance can graduate into org memory:
- Confirmed findings require at least 3 verified samples and a verified rate of at least 50%.
- Accepted suppressions require the same stable key to be manually suppressed at least twice.
- Demotion hints reuse `MIN_SAMPLES_FOR_DEMOTE` from `feedback_collector.py` to avoid single-sample overfit.
By default the compiler adds `.vuln-scout/org-memory/` to the target repository `.gitignore`. Use `--allow-commit` only when the repository owner has decided the generated memory is safe to commit.
Read more
name: org-memory-compile description: "[experimental] Compile human-reviewed scan history into local organization memory" argument-hint: "[--privacy open|hashed|strict] [--dry-run] [--allow-commit] [--force]" allowed-tools: - Bash - Read - Write - Glob - Grep
Compile Org Memory
Compile repeated human-reviewed verdicts into `.vuln-scout/org-memory/` so future scans can reuse organization-specific suppressions, confirmed patterns, and reviewer heuristics.
Usage
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/org_memory_compiler.py" \
--project-root . \
--privacy hashedpython3 "${CLAUDE_PLUGIN_ROOT}/scripts/org_memory_compiler.py" \
--project-root . \
--privacy strict \
--dry-runInputs
- `.claude/scan-history/*.json`
- `.claude/rule-stats.json`
- `.claude/review-ledger.json`
- Human-reviewed findings where `verdict` is `verified` or `false_positive`
Outputs
- `.vuln-scout/org-memory/accepted-suppressions.yaml`
- `.vuln-scout/org-memory/confirmed-findings.yaml`
- `.vuln-scout/org-memory/custom-rules/semgrep/*.yaml`
- `.vuln-scout/org-memory/review-patterns.yaml`
- `.vuln-scout/org-memory/manifest.json`
Flags
| Flag | Effect | |------|--------| | `--privacy open` | Store paths and excerpts verbatim | | `--privacy hashed` | Store SHA-256 hashes for paths and excerpts | | `--privacy strict` | Store only rule IDs, verdict counts, and CWE metadata | | `--dry-run` | Print proposed memory without writing files | | `--force` | Allow overwriting strict memory with open privacy | | `--allow-commit` | Do not add `.vuln-scout/org-memory/` to `.gitignore` |
Review Policy
Only human-reviewed provenance can graduate into org memory:
- Confirmed findings require at least 3 verified samples and a verified rate of at least 50%.
- Accepted suppressions require the same stable key to be manually suppressed at least twice.
- Demotion hints reuse `MIN_SAMPLES_FOR_DEMOTE` from `feedback_collector.py` to avoid single-sample overfit.
By default the compiler adds `.vuln-scout/org-memory/` to the target repository `.gitignore`. Use `--allow-commit` only when the repository owner has decided the generated memory is safe to commit.
AI-powered whitebox penetration testing plugin for Claude Code. 9 languages, 22 skills, 7 autonomous agents. STRIDE threat modeling, OWASP 2025 coverage, polyglot monorepo support.
Repo: allsmog/vuln-scout
Other commands on vuln-scout.
- /auto-fix
[experimental] Auto-remediate verified findings by generating patches and optionally creating a PR
Open command - /create-rule
[experimental] Create a custom Semgrep detection rule from a confirmed vulnerability pattern
Open command - /diff
[stable] Compare security posture between two git refs to find new/fixed vulnerabilities and track regression
Open command - /full-audit
[stable] End-to-end security audit with hotspot-aware framework pivots, shared findings.json schema, and CI-friendly workflow flags
Open command - /mobile-audit
[beta] Audit a decompiled Android target — scans jadx_out/sources + apktool_out together and merges findings
Open command - /mutate
[experimental] Security mutation testing -- weaken security controls and check if the scanner detects the resulting vulnerability
Open command

