master
7b5bdcf4d3
- Introduced AGENTS.md, README.md, TASKS.md, and implementation_plan.md for Vexer, detailing mission, responsibilities, key components, and operational notes. - Established similar documentation structure for Vulnerability Explorer and Zastava modules, including their respective workflows, integrations, and observability notes. - Created risk scoring profiles documentation outlining the core workflow, factor model, governance, and deliverables. - Ensured all modules adhere to the Aggregation-Only Contract and maintain determinism and provenance in outputs.
101 lines
4.8 KiB
Markdown
101 lines
4.8 KiB
Markdown
# Advisory AI architecture
|
|
|
|
> Captures the retrieval, guardrail, and inference packaging requirements defined in the Advisory AI implementation plan and related module guides.
|
|
|
|
## 1) Goals
|
|
|
|
- Summarise advisories/VEX evidence into operator-ready briefs with citations.
|
|
- Explain conflicting statements with provenance and trust weights (using VEX Lens & Excititor data).
|
|
- Suggest remediation plans aligned with Offline Kit deployment models and scheduler follow-ups.
|
|
- Operate deterministically where possible; cache generated artefacts with digests for audit.
|
|
|
|
## 2) Pipeline overview
|
|
|
|
```
|
|
+---------------------+
|
|
Concelier/VEX Lens | Evidence Retriever |
|
|
Policy Engine ----> | (vector + keyword) | ---> Context Pack (JSON)
|
|
Zastava runtime +---------------------+
|
|
|
|
|
v
|
|
+-------------+
|
|
| Prompt |
|
|
| Assembler |
|
|
+-------------+
|
|
|
|
|
v
|
|
+-------------+
|
|
| Guarded LLM |
|
|
| (local/host)|
|
|
+-------------+
|
|
|
|
|
v
|
|
+-----------------+
|
|
| Citation & |
|
|
| Validation |
|
|
+-----------------+
|
|
|
|
|
v
|
|
+----------------+
|
|
| Output cache |
|
|
| (hash, bundle) |
|
|
+----------------+
|
|
```
|
|
|
|
## 3) Retrieval & context
|
|
|
|
- Hybrid search: vector embeddings (SBERT-compatible) + keyword filters for advisory IDs, PURLs, CVEs.
|
|
- Context packs include:
|
|
- Advisory raw excerpts with highlighted sections and source URLs.
|
|
- VEX statements (normalized tuples + trust metadata).
|
|
- Policy explain traces for the affected finding.
|
|
- Runtime/impact hints from Zastava (exposure, entrypoints).
|
|
- Export-ready remediation data (fixed versions, patches).
|
|
|
|
All context references include `content_hash` and `source_id` enabling verifiable citations.
|
|
|
|
## 4) Guardrails
|
|
|
|
- Prompt templates enforce structure: summary, conflicts, remediation, references.
|
|
- Response validator ensures:
|
|
- No hallucinated advisories (every fact must map to input context).
|
|
- Citations follow `[n]` indexing referencing actual sources.
|
|
- Remediation suggestions only cite policy-approved sources (fixed versions, vendor hotfixes).
|
|
- Moderation/PII filters prevent leaking secrets; responses failing validation are rejected and logged.
|
|
|
|
## 5) Output persistence
|
|
|
|
- Cached artefacts stored in `advisory_ai_outputs` with fields:
|
|
- `output_hash` (sha256 of JSON response).
|
|
- `input_digest` (hash of context pack).
|
|
- `summary`, `conflicts`, `remediation`, `citations`.
|
|
- `generated_at`, `model_id`, `profile` (Sovereign/FIPS etc.).
|
|
- `signatures` (optional DSSE if run in deterministic mode).
|
|
- Offline bundle format contains `summary.md`, `citations.json`, `context_manifest.json`, `signatures/`.
|
|
|
|
## 6) Profiles & sovereignty
|
|
|
|
- **Profiles:** `default`, `fips-local` (FIPS-compliant local model), `gost-local`, `cloud-openai` (optional, disabled by default). Each profile defines allowed models, key management, and telemetry endpoints.
|
|
- **CryptoProfile/RootPack integration:** generated artefacts can be signed using configured CryptoProfile to satisfy procurement/trust requirements.
|
|
|
|
## 7) APIs
|
|
|
|
- `POST /v1/advisory-ai/summaries` — generate (or retrieve cached) summary for `{advisoryKey, artifactId, policyVersion}`.
|
|
- `POST /v1/advisory-ai/conflicts` — explain conflicting VEX statements with trust ranking.
|
|
- `POST /v1/advisory-ai/remediation` — fetch remediation plan with target fix versions, prerequisites, verification steps.
|
|
- `GET /v1/advisory-ai/outputs/{hash}` — retrieve cached artefact (used by CLI/Console/Export Center).
|
|
|
|
All endpoints accept `profile` parameter (default `fips-local`) and return `output_hash`, `input_digest`, and `citations` for verification.
|
|
|
|
## 8) Observability
|
|
|
|
- Metrics: `advisory_ai_requests_total{profile,type}`, `advisory_ai_latency_seconds`, `advisory_ai_validation_failures_total`.
|
|
- Logs: include `output_hash`, `input_digest`, `profile`, `model_id`, `tenant`, `artifacts`. Sensitive context is not logged.
|
|
- Traces: spans for retrieval, prompt assembly, model inference, validation, cache write.
|
|
|
|
## 9) Operational controls
|
|
|
|
- Feature flags per tenant (`ai.summary.enabled`, `ai.remediation.enabled`).
|
|
- Rate limits (per tenant, per profile) enforced by Orchestrator to prevent runaway usage.
|
|
- Offline/air-gapped deployments run local models packaged with Offline Kit; model weights validated via manifest digests.
|