po4yka/llm-wiki-skills / skills/llm-wiki-threat-model
llm-wiki-threat-model
VerifiedBuild a formal security architecture threat model for an LLM-Wiki system. Use when the user asks for STRIDE, LINDDUN, PASTA, data-flow diagrams, trust boundaries, attack-surface mapping, abuse cases, risk matrices, red-team scenarios, or control baselines; route existing-config safety reviews to llm-wiki-security-review.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.0 KB |
| references/docs/19-security-threat-model.md | markdown | 23.0 KB |
| references/policies/redaction-retention-policy.md | markdown | 3.7 KB |
| references/policies/review-incident-response.md | markdown | 3.1 KB |
| references/templates/llm-wiki-security.github-actions.yml | config | 2.9 KB |
| references/templates/mcp-security-profile.yaml | config | 3.6 KB |
| references/templates/promptfoo-llm-wiki-redteam.yaml | config | 3.8 KB |
| references/templates/security-scorecard.yaml | config | 4.0 KB |
references/policies/redaction-retention-policy.md
View on GitHub# Redaction and retention policy for LLM-Wiki
> Status: template
> Scope: raw sources, wiki pages, indexes, traces, logs, eval artifacts and exports.
## Purpose
This policy defines what may be captured, indexed, retrieved, traced, exported and retained in an LLM-Wiki system.
## Data classes
| Class | Examples | Default handling |
| --- | --- | --- |
| public | Published docs, public repo content, public web pages. | May be indexed and exported when provenance is preserved. |
| internal | Team docs, internal project notes, non-public architecture. | May be indexed for authorized users; export requires internal profile. |
| sensitive | Private operational data, customer-related material, confidential business context. | Requires classification, filters and redaction before external processing or export. |
| regulated | Data governed by legal, contractual or compliance rules. | Do not process externally or export without explicit policy approval. |
| unknown | Unclassified captures or raw imports. | Treat as sensitive until classified. |
## Capture and raw storage
- Raw sources are stored only when there is a clear purpose.
- Raw sources must have manifests with source ID, origin, hash, capture time, owner and sensitivity.
- Unknown or sensitive raw sources are excluded from production retrieval until classified.
- Parser output must be scanned before indexing or export.
- Raw source deletion or mutation through MCP/API tools is disabled by default.
## Indexing and retrieval
- Production retrieval must filter by tenant, sensitivity, publication state and review state.
- Draft, rejected, stale and quarantined content is excluded from production retrieval unless explicitly requested for review.
- Vector, FTS and graph indexes inherit the highest sensitivity of their source chunks.
- Indexes are rebuildable and must not become the only durable copy of sensitive content.
## Redaction
Before external processing, tracing or export, scan for sensitive identifiers, confidential business context, private user content and restricted operational details.
Redaction reports must include:
```yaml
redaction_report:
generated_at: "YYYY-MM-DDTHH:MM:SSZ"
input_paths: []
output_paths: []
detectors: []
findings_count: 0
blocked: true|false
reviewer: ""
```
## Traces and logs
- Traces and logs are sensitive stores.
- Do not publicly share traces containing prompts, retrieved snippets or private source material.
- Prefer hashes, source IDs and page IDs over raw source text in audit logs.
- Keep retention periods short for raw prompts and retrieved snippets.
- Audit logs for sensitive reads, proposals and admin tools should be retained according to team policy.
## Exports
- Public exports are allowlist-based.
- `raw/` is excluded from public exports by default.
- `llms.txt`, JSONL, JSON-LD, GraphML and static-site exports require redaction checks.
- Export manifests must list included spaces, excluded paths, sensitivity classes, redaction status and reviewer.
## Retention
| Artifact | Default retention |
| --- | --- |
| Raw public sources | Keep while useful and cited. |
| Raw internal/sensitive sources | Review retention periodically; minimize where possible. |
| Parser temp output | Delete after successful manifest/index/page creation. |
| Traces with snippets | Short retention unless explicitly approved. |
| Audit logs | Retain according to security/compliance policy. |
| Public export artifacts | Retain with release/archive manifest. |
| Red-team fixtures | Keep isolated; never mix with production sources. |
## Exceptions
Every exception must include:
```yaml
exception:
owner: ""
reason: ""
affected_paths: []
data_class: sensitive|regulated|unknown
compensating_controls: []
expires_at: "YYYY-MM-DD"
approver: ""
```
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |