Open-source AI Security Operations Center: alert fusion, LLM-agent triage, MITRE ATT&CK investigation, and a replayable decision ledger for every agent step. Self-hostable, runs with no API keys, MIT licensed. Ships an MCP server for Claude, Cursor and Continue.
Python agentic-aiai-securityalert-triageblue-teamcybersecuritydetection-engineeringincident-responselanggraphllm-agentsmcpmitre-attackmodel-context-protocolpurple-teamsecurity-operationsself-hostedsiemsoarsocthreat-detectionthreat-intelligence
Commits on main
fix(api): reconcile guards with the reopened case surface
46e6867 Alexandru Mateescu committed 1 month ago
chore: keep vendor product names inside integration surfaces
5445ca9 Alexandru Mateescu committed 1 month ago
feat(dashboard): global time window drives the overview; notifications navigate
eb2c580 Alexandru Mateescu committed 1 month ago
test(api): MSSP isolation fixtures seed the live case table
d24c72c Alexandru Mateescu committed 1 month ago
fix(api): honest dashboard metrics from the live case table
4bbfbc1 Alexandru Mateescu committed 1 month ago
fix(web): alert filter chips render canonical labels, not raw lowercase
46fff44 Alexandru Mateescu committed 1 month ago
fix(api,web): attack-chain 404, reopen endpoint, investigate auto-advances status
9f3164d Alexandru Mateescu committed 1 month ago
fix(agents,api): send real alert payload to investigation; stop fabricated evidence
128ebb0 Alexandru Mateescu committed 1 month ago
fix(alerts): Raw tab shows original event + Wazuh locator
c5a9332 Alexandru Mateescu committed 1 month ago
docs: maintenance index, fix log, ops runbook, known gaps (Oct 2026 session)
c5f8f60 Alexandru Mateescu committed 1 month ago
chore(deploy): ollama GPU reservation + agents API key wiring (sanitized)
03b4a65 Alexandru Mateescu committed 1 month ago
fix(web): investigation tab survives refresh; responder cards render; overview timeline populated
d95e27d Alexandru Mateescu committed 1 month ago
fix(agents): coerce nested LLM JSON to prose in ForensicFindings
14615ec Alexandru Mateescu committed 1 month ago
fix(api): resolve real permission set when dispatching approval decisions
7db50b4 Alexandru Mateescu committed 1 month ago
fix(alerts): status follows agent verdicts; triaged visible in queue; release-assignee works
5651253 Alexandru Mateescu committed 1 month ago
fix(api): case status ladder, agents-proxy auth, ledger-backed investigations list
9a59f24 Alexandru Mateescu committed 1 month ago
fix(agents): forward aisoc_ API key on all core-API calls
da394c0 Alexandru Mateescu committed 1 month ago
fix(api): attach service credentials to agents-service proxies
0938dc6 Alexandru Mateescu committed 1 month ago
docs: the model-placement and metrics work, and three pages that were wrong (#1125)
3795bac Beenu Arora committed 1 month ago
fix(metrics): count open work, score what the agent says, and reconcile case status (#1124)
f08b859 Beenu Arora committed 1 month ago
feat(model): let users choose where the model runs (#1123)
21fe39f Beenu Arora committed 1 month ago
fix(security): two reported advisories — WebSocket auth and SSO account binding (#1122)
709ea6c Beenu Arora committed 1 month ago
fix(wave5): hunting, intel, sandbox and operations wiring (5.1-5.8) (#1120)
92aef46 Beenu Arora committed 1 month ago
fix(wave6): retract what is not built, and remove routes that cannot run (#1119)
7657048 Beenu Arora committed 1 month ago
fix(wave 3, partial): replay and shadow mode measure the real run (#1118)
d25d655 Beenu Arora committed 1 month ago
fix(wave 2): the MCP client works over a real connection (#1117)
3236335 Beenu Arora committed 1 month ago
fix(wave 1): security and tenancy defects that every green CI run missed (#1116)
614e64b Beenu Arora committed 1 month ago
fix(security): measure a grant against the authority its caller was admitted on (#1114)
b1c6925 Beenu Arora committed 1 month ago