F
fabriziosalmi/security-audit-cli
Deterministic 6-phase security audit CLI, MCP server, and VS Code extension (1:1:1 parity, *os.Root…
updated 2026-11-17
View on GitHub ↗Link checked 2026-11-17 · HTTP 200 · opens the vendor's own site
After install: how to tell it works
- What you see after install — The command exits 0 and the terminal shows no "command not found" or "permission denied".
- One-line self-check — Run the smallest usage example from that repository's README. If the output matches what the README describes, it is working.
- If that does not work — This project has no troubleshooting notes. Search its issues for "install".
The install command comes from the repository; these checks describe observable results only and do not replace the project's own docs.
- Star count
- 0 ★ · fabriziosalmi/security-audit-cli · read from the GitHub search API · (checked 2026-11-19 20:00:48)
- Source
- read directly from its GitHub page · type: open-source project
- Harvest history
- first seen 2026-11-16 · last seen 2026-11-20 · seen 5× · listed by 4 sources
- Score
- score 403 · 4 sources · repo
- Install
- see Install & config above
FAQ (answers come from our records)
Who is it for?
We filed it under these tags: Open Source. The tags come from the entry itself, not a taxonomy we imposed.
Who maintains it?
repository
fabriziosalmi/security-audit-cli;first catalogued 2026-11-16;last seen 2026-11-20。There is no official certification or quality rating here — we record sources and dates only.README excerpt (English, untranslated) · Read the full README on GitHub →
# sec-audit
Excerpted from the repository README (abridged)
sec-audit is a deterministic Go CLI and Model Context Protocol (MCP stdio JSON-RPC 2.0) server that enforces a 6-phase adversarial security audit state machine over source repositories.
Built on the methodology and embedded domain playbooks of cloudflare/security-audit-skill, sec-audit replaces prompt-only instructions with hard programmatic gates: every coverage unit, source citation, local verification check, adversarial role transition, and final report artifact is verified on disk before the state machine advances.
One binary. Go 1.24+ standard library only. Zero external dependencies. Strict 1:1:1 parity across CLI subcommands, MCP tools, and the optional VS Code / Cursor / Antigravity extension (editors/vscode/) via a single generic command registry.
## Table of Contents
Why sec-audit
6-Phase State Machine
1:1:1 CLI Subcommands, MCP Tools, and VS Code Commands
Deterministic Security and Reproducibility Invariants
Installation
MCP Server Configuration
VS Code / Cursor / Antigravity Extension
End-to-End CLI Walkthrough
Canonical Output Artifacts
Measured Performance and Benchmarks
Repository Architecture
License
## Why sec-audit
Multi-agent security audits degrade when orchestrators rely on unverified LLM prose:
1. Phantom citations: Agents cite non-existent files, symlinks escaping the repository, or out-of-range line numbers.
2. Self-grading bias: The same agent that proposes a hypothesis also confirms it or rewrites the final report record.
3. Silent coverage gaps: Attack-class playbooks or repository entrypoints are silently skipped without explicit exclusion rationale.
4. Parser differentials: Duplicate JSON keys ({"verdict":"rejected","verdict":"confirmed"}) or deeply nested payloads slip past standard unmarshalers.
5. Non-reproducible outputs: Unordered map iteration produces different prompts, ledgers, or reports across runs.
sec-audit eliminates these failure modes deterministically in code:
Every CLI subcommand (sec-audit ) and MCP tool (sec_audit_ ) is bound through a single generic [Bind[Req, Resp]](internal/registry/registry.go) registration, guaranteeing 1:1 schema and execution parity by construction.
All 11 attack-class catalogs (ATTACK-CLASSES.md + 10 domain companions) and report-schema.json are embedded in the binary via //go:embed (internal/spec).
Phase 6 report rendering is blocked until Validate confirms 0 ledger errors, 0 finding errors, 100% cross-file fingerprint linkage, and adversarial role separation (hunter != verifier != record_verifier).
## 6-Phase State Machine
mermaid
flowchart TD
P1A["Phase 1a: init\n(Repo scan, budget gate, prior carryover)"] --> P1B["Phase 1b: architecture\n(Write architecture.md P1C["Phase 1c: ledger-seed\n(Derive RFC 3986 + NFC coverage_id, verify paths)"]
P1C --> P2A["Phase 2: agent-assign (hunter)\n(Compile verbatim playbook prompt, isolate scratch)"]
P2A --> P2B["Phase 2: artifact-promote + hunter-submit\n(11-step os.Root promote, verify reviewed_paths)"]
P2B --> P2C["Phase 2: critic-submit (post_wave / final_clean)\n(Archive reopened attempts[], add missing units)"]
P2C -->|"More waves required"| P2A
P2C -->|"final_clean stop=true"| P3["Phase 3: verifier-submit\n(Enforce verifier != hunter, verify trace on disk)"]
Built on the methodology and embedded domain playbooks of cloudflare/security-audit-skill, sec-audit replaces prompt-only instructions with hard programmatic gates: every coverage unit, source citation, local verification check, adversarial role transition, and final report artifact is verified on disk before the state machine advances.
One binary. Go 1.24+ standard library only. Zero external dependencies. Strict 1:1:1 parity across CLI subcommands, MCP tools, and the optional VS Code / Cursor / Antigravity extension (editors/vscode/) via a single generic command registry.
## Table of Contents
Why sec-audit
6-Phase State Machine
1:1:1 CLI Subcommands, MCP Tools, and VS Code Commands
Deterministic Security and Reproducibility Invariants
Installation
MCP Server Configuration
VS Code / Cursor / Antigravity Extension
End-to-End CLI Walkthrough
Canonical Output Artifacts
Measured Performance and Benchmarks
Repository Architecture
License
## Why sec-audit
Multi-agent security audits degrade when orchestrators rely on unverified LLM prose:
1. Phantom citations: Agents cite non-existent files, symlinks escaping the repository, or out-of-range line numbers.
2. Self-grading bias: The same agent that proposes a hypothesis also confirms it or rewrites the final report record.
3. Silent coverage gaps: Attack-class playbooks or repository entrypoints are silently skipped without explicit exclusion rationale.
4. Parser differentials: Duplicate JSON keys ({"verdict":"rejected","verdict":"confirmed"}) or deeply nested payloads slip past standard unmarshalers.
5. Non-reproducible outputs: Unordered map iteration produces different prompts, ledgers, or reports across runs.
sec-audit eliminates these failure modes deterministically in code:
Every CLI subcommand (sec-audit ) and MCP tool (sec_audit_ ) is bound through a single generic [Bind[Req, Resp]](internal/registry/registry.go) registration, guaranteeing 1:1 schema and execution parity by construction.
All 11 attack-class catalogs (ATTACK-CLASSES.md + 10 domain companions) and report-schema.json are embedded in the binary via //go:embed (internal/spec).
Phase 6 report rendering is blocked until Validate confirms 0 ledger errors, 0 finding errors, 100% cross-file fingerprint linkage, and adversarial role separation (hunter != verifier != record_verifier).
## 6-Phase State Machine
mermaid
flowchart TD
P1A["Phase 1a: init\n(Repo scan, budget gate, prior carryover)"] --> P1B["Phase 1b: architecture\n(Write architecture.md P1C["Phase 1c: ledger-seed\n(Derive RFC 3986 + NFC coverage_id, verify paths)"]
P1C --> P2A["Phase 2: agent-assign (hunter)\n(Compile verbatim playbook prompt, isolate scratch)"]
P2A --> P2B["Phase 2: artifact-promote + hunter-submit\n(11-step os.Root promote, verify reviewed_paths)"]
P2B --> P2C["Phase 2: critic-submit (post_wave / final_clean)\n(Archive reopened attempts[], add missing units)"]
P2C -->|"More waves required"| P2A
P2C -->|"final_clean stop=true"| P3["Phase 3: verifier-submit\n(Enforce verifier != hunter, verify trace on disk)"]
Listed badge (optional)
Copy this line into the README if you want the listing mark. Optional, and it does not affect being listed.
[](https://anclum.com/en/s/fabriziosalmi-security-audit-cli/)
- Source
- github · https://github.com/fabriziosalmi/security-audit-cli
- Browse by field
- Open source
- Listed by
- 4 independent sources
- added
- today
More entries
Filed under
Every tag opens the full list for that filter. Counts are the live catalogue figures.