devatsecure/Argus Security

The All-Seeing AI Security Platform: Multi-agent security analysis with 5 specialized AI personas, 9 security scanners, Claude/OpenAI triage, and intelligent false positive reduction

View on GitHub

Trust Signals

Scorecard Score
not yet scored
Maintenance Recency
Activelast commit Mar 28, 2026
License
MIT

Pinned Snippet

workflow.ymlSHA-pinned
uses: devatsecure/argus-security@7a6b42c8ab79016743c5115c0f360720ebaeb05b # v3.1.0

tags can be moved; commit SHAs can't. why a SHA?

namedescriptionrequireddefault
pipeline-modePipeline mode: "fast" uses run_ai_audit.py (AI code review, 2-3 LLM calls), "full" uses hybrid_analyzer.py (6-phase pipeline with Semgrep+Trivy+Checkov+TruffleHog+AI enrichment)nofast
review-typeType of review to run (audit, security, review) — used in fast modenoaudit
project-pathPath to the project to reviewno.
project-typeProject type (backend-api, dashboard-ui, data-pipeline, infrastructure, or auto) — used in fast modenoauto
fail-on-blockersFail the workflow if merge blockers are foundnotrue
fail-onGranular fail conditions (e.g., "security:high,security:critical,test:critical")no""
comment-on-prComment on PR with review resultsnotrue
upload-reportsUpload review reports as artifactsnotrue
ai-providerAI provider to use: anthropic, openai, openrouter, ollama, or autonoauto
anthropic-api-keyAnthropic API key for Claude AI analysisno""
openai-api-keyOpenAI API key for GPT-4 analysisno""
openrouter-api-keyOpenRouter API key for DeepSeek/Qwen/200+ modelsno""
openrouter-modelOpenRouter model ID (e.g., deepseek/deepseek-v3.2, qwen/qwen3-235b-a22b)nodeepseek/deepseek-v3.2
semgrep-enabledEnable Semgrep SAST scanning with 2,000+ security rulesnotrue
ollama-endpointOllama endpoint URL for local LLM (e.g., http://localhost:11434)no""
modelAI model to use (e.g., claude-sonnet-4, gpt-4, llama3)noauto
multi-agent-modeEnable multi-agent mode: single, sequential, or parallelnosingle
enable-multi-agentEnable specialized agent personas (SecretHunter, ArchitectureReviewer, ExploitAssessor, FalsePositiveFilter, ThreatModeler) for collaborative security analysisnotrue
enable-spontaneous-discoveryEnable agents to find security issues beyond scanner rules (15-20% more findings): missing controls, architectural flaws, implicit trust issues, config mistakes, supply chain risksnotrue
enable-collaborative-reasoningEnable multi-agent collaboration and discussion for consensus-based decisions (30-40% fewer false positives, but 40% higher cost)nofalse
enable-api-securityEnable OWASP API Top 10 testing for REST/GraphQL/gRPC endpoints (BOLA, broken auth, SSRF, misconfigurations)notrue
enable-dastEnable DAST scanning with Nuclei (4000+ templates) - Requires dast-target-url parameternofalse
dast-target-urlTarget URL for DAST scanning (e.g., https://staging.example.com)no""
enable-sandboxEnable Docker sandbox validation for exploit verification (Phase 4) — requires Docker, full mode onlynofalse
enable-proof-by-exploitationEnable LLM-powered proof-of-concept exploit generation + sandbox validation — requires Docker + LLM credits, full mode onlynofalse
enable-supply-chainEnable supply chain attack detection (typosquatting, malicious packages, compromised dependencies)notrue
enable-fuzzingEnable AI-guided fuzzing with 60+ payloads for APIs, functions, and file parsersnofalse
fuzzing-durationFuzzing duration in seconds (default: 300)no300
enable-threat-intelEnable threat intelligence enrichment (CISA KEV, EPSS, NVD, exploit databases, real-time threat context)notrue
enable-remediationEnable AI-powered vulnerability fix generation with code patches and testing guidancenotrue
enable-runtime-securityEnable container runtime security monitoring with Falco (requires Docker)nofalse
runtime-monitoring-durationRuntime monitoring duration in seconds (default: 60)no60
enable-regression-testingEnable security regression test generation to ensure fixed vulnerabilities stay fixednotrue
enable-irisEnable IRIS semantic analysis for research-proven 2x improvement in vulnerability detection (arXiv 2405.17238)notrue
severity-filterComma-separated severity levels to include in reports (e.g., critical,high,medium). Empty means all severities.no""
enable-exploit-analysis✅ PRODUCTION: Enable exploit chain analysis and exploitability assessment (Aardvark mode) - Working in sequential modenotrue
generate-security-tests✅ PRODUCTION: Auto-generate security tests for discovered vulnerabilities - Working in sequential modenotrue
exploitability-thresholdBlock merge if exploitability is at this level or higher (trivial, moderate, complex, theoretical, none)notrivial
only-changedOnly analyze changed files (PR mode)nofalse
include-pathsGlob patterns to include (comma-separated)no""
exclude-pathsGlob patterns to exclude (comma-separated)no.github/**,node_modules/**,*.lock,package-lock.json
max-file-sizeMax file size in bytesno50000
max-filesMax files to analyzeno50
max-tokensMax tokens per LLM callno8000
cost-limitMax cost in USD per runno1.0
deep-analysis-modePhase 2.7 Deep Analysis mode (off, semantic-only, conservative, full)nooff
max-files-deep-analysisMaximum files for Phase 2.7 Deep Analysisno50
deep-analysis-cost-ceilingCost ceiling in USD for Phase 2.7no5.0
deep-analysis-timeoutTimeout in seconds for Phase 2.7no300
benchmarkEnable benchmark reportingnofalse
namedescription
review-completedWhether the review completed successfully
blockers-foundNumber of merge blockers found
suggestions-foundNumber of suggestions found
report-pathPath to the generated report
sarif-pathPath to SARIF file for Code Scanning
json-pathPath to structured JSON results
cost-estimateEstimated cost in USD
files-analyzedNumber of files analyzed
duration-secondsAnalysis duration in seconds
exploitability-trivialNumber of trivially exploitable vulnerabilities found
exploitability-moderateNumber of moderately exploitable vulnerabilities found
exploitability-complexNumber of complex exploitability vulnerabilities found
exploit-chains-foundNumber of exploit chains identified
tests-generatedNumber of security test files generated
pipeline-mode-usedWhich pipeline mode was used (fast or full)
total-findingsTotal findings from full pipeline scan
scanners-usedComma-separated list of scanners that ran
scan-duration-secondsFull pipeline scan duration in seconds