Major additions: - Community Next.js app (port 18187) for browsing claims with API docs - stemedb-chaos crate: Fault injection, chaos testing, CRDT properties - Latent ingestion system: Reddit/FDA ingesters with ADK-Go agents - Disputed claims handling: Manual review workflows and validation - Aphoria security scanner: New extractors (SQL injection, command injection, weak crypto, TLS version), policy-based ignores, UAT reports - Docker infrastructure: Dockerfile, docker-compose.yml for full stack - VulnBank demo: Intentionally vulnerable multi-language test corpus SDK & API enhancements: - Source registry handlers for tracking data provenance - Metrics endpoint - Skeptic filtering improvements Code quality: - Split 14 large files (>500 lines) into focused modules - All files now under 500-line limit per project guidelines Documentation: - Chaos testing guide, circuit breakers, observability docs - Phase 7 UAT documentation updates - Martin Kleppmann technical writer agent Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
8.8 KiB
Phase 7 UAT: The Shield
Status: Ready for Testing Target Date: 2026-02-03 Confidence: High (7A, 7B complete; 7C core complete)
Summary
Phase 7 (The Shield) defends against spam, Sybil attacks, and knowledge poisoning. This UAT validates the trust-at-scale infrastructure for opening Episteme to millions of agents.
Scope:
- 7A Admission Control: PoW-based spam protection, trust tiers, graduated quotas
- 7B EigenTrust: Sybil-resistant global trust propagation
- 7C Content Defense: Quality scoring, MinHash/LSH similarity, quarantine store, admin API
- 7D Circuit Breakers: Per-agent banning with Closed/Open/HalfOpen state machine
Test Coverage (Verified)
| Area | Tests | Status |
|---|---|---|
| Circuit Breakers (7D) | 25 | PASS |
| Trust Graph Store (7B) | 23 | PASS |
| Trust Rank Store | 22 | PASS |
| PoW types (7A) | 19 | PASS |
| Domain Trust Store (7B) | 18 | PASS |
| Admission Store (7A) | 16 | PASS |
| Content Defense (quality) (7C) | 13 | PASS |
| Similarity Index (MinHash/LSH) (7C) | 12 | PASS |
| Quarantine Store (7C) | 9 | PASS |
| Trust Tier types (7A) | 8 | PASS |
| API Admission integration | 6 | PASS |
| Content Defense Layer (7C) | 5 | PASS |
| Total Phase 7 | 176 | ALL PASS |
Realistic Usage Scenarios
Scenario 1: New Agent Onboarding
Goal: Verify graduated difficulty protects against spam bots while not blocking legitimate agents.
# 1. New agent with no history should require PoW
curl -X GET http://localhost:18180/v1/admission/status \
-H "X-Agent-Id: 0000000000000000000000000000000000000000000000000000000000000001"
# Expected: 200 with pow_required: true, difficulty: 16
# 2. Submit first assertions with PoW proof
# Agent must solve: BLAKE3(nonce || agent_id || timestamp) has 16 leading zero bits
# This takes ~16 seconds on average
# 3. After 10 assertions, difficulty drops to 1 bit (trivial)
# 4. After 50 assertions OR trust > 0.6, PoW exempt
Acceptance Criteria:
- New agents see
pow_required: true,difficulty: 16 - HTTP 428 returned when PoW missing/invalid
- Difficulty graduates: 16 bits (1-10) → 1 bit (11-50) → 0 (51+)
- Trusted agents (>0.6) are exempt regardless of assertion count
Scenario 2: Trust Tier Quotas
Goal: Verify rate limiting scales with trust level.
| Tier | Trust Range | Quota Multiplier | Hourly Limit |
|---|---|---|---|
| Untrusted | 0.0-0.3 | 0.1x | 1,000/hr |
| Limited | 0.3-0.5 | 0.5x | 5,000/hr |
| Verified | 0.5-0.7 | 1.0x | 10,000/hr |
| Trusted | 0.7-0.9 | 2.0x | 20,000/hr |
| Authority | 0.9-1.0 | 10.0x | 100,000/hr |
Acceptance Criteria:
- Quota headers present in responses (
X-RateLimit-*) - Untrusted agents limited to 0.1x base quota
- Authority agents get 10x quota
- HTTP 429 returned when quota exceeded
Scenario 3: EigenTrust Sybil Resistance
Goal: Verify isolated trust rings get near-zero global trust.
Legitimate Network: Sybil Ring:
Seed ─────> A X ──> Y
│ │ │ │
v v v v
B ──────> C Z <── W
Acceptance Criteria:
- Seed-connected agents (A, B, C) accumulate positive global trust
- Isolated ring (X, Y, Z, W) converges to near-zero trust
- Power iteration converges in <100 iterations (ε = 1e-4)
- Domain-specific trust factors applied correctly
Scenario 4: Content Quality Filtering
Goal: Verify spam/noise detection without blocking legitimate content.
| Content Type | Expected Quality | Should Quarantine? |
|---|---|---|
| Normal assertion: "Aspirin:treats:Headache" | >0.6 | No |
| Low entropy: "aaaa:bbbb:cccc" | <0.4 | Yes |
| Structured data with JSON | >0.7 (bonus) | No |
| Untrusted agent + high confidence | <0.5 (penalty) | Yes |
Acceptance Criteria:
- Shannon entropy check flags random noise (< 1.5 bits/char)
- Minimum subject/predicate length enforced (default 3 chars)
- Structured data (JSON, URLs, dates) gets +0.1 bonus
- Untrusted + high confidence gets -0.5 penalty
- Quality < 0.4 triggers quarantine
Scenario 5: Quarantine Admin Workflow
Goal: Verify suspicious content can be reviewed and processed.
# 1. List pending quarantine events
curl http://localhost:18180/v1/admin/quarantine?limit=20
# 2. Review specific event
curl http://localhost:18180/v1/admin/quarantine/{hash}
# 3. Approve or reject
curl -X POST http://localhost:18180/v1/admin/quarantine/{hash}/approve
curl -X POST http://localhost:18180/v1/admin/quarantine/{hash}/reject
Acceptance Criteria:
GET /v1/admin/quarantinelists pending events with reasonsGET /v1/admin/quarantine/{hash}returns full assertion bytesPOST .../approvemoves assertion to main indexPOST .../rejectmarks as reviewed but keeps quarantined- Quarantine reasons clearly indicate why flagged
Integration Points to Verify
-
Ingestion Pipeline Integration
- Content defense layer called before indexing
- Quarantine bypasses normal index path
- Bloom filter restored on restart
-
Trust Store Interplay
- EigenTrust feeds into TrustTier calculation
- Domain trust factors into Authority lens weights
- Trust decay applies to computed scores
-
API Middleware Chain
- AdmissionLayer checks PoW before rate limiting
- MeterLayer applies tier-based quotas
- Headers reflect current trust state
Scenario 6: Near-Duplicate Detection (MinHash/LSH)
Goal: Verify similar content is flagged without blocking unique assertions.
| Content Pair | Jaccard Similarity | Expected |
|---|---|---|
| "Aspirin:treats:Headache" vs same | 1.0 | Duplicate |
| "Aspirin:treats:Headache" vs "Aspirin:treats:Migraine" | ~0.7 | Unique |
| "Aspirin treats headaches" vs "Aspirin:treats:Headache" | ~0.85 | Unique |
| "Asprin:treats:Headach" (typos) vs "Aspirin:treats:Headache" | ~0.92 | Duplicate |
Acceptance Criteria:
- Bloom filter provides fast "definitely not seen" path
- MinHash signatures (k=128) computed correctly
- LSH bands (16 bands × 8 rows) enable efficient lookup
- Jaccard threshold of 0.9 correctly identifies near-duplicates
- Unique content with similar structure passes through
Scenario 7: Circuit Breaker State Machine (7D)
Goal: Verify misbehaving agents are temporarily banned with recovery path.
State Flow:
CLOSED (normal) --[5 failures]--> OPEN (banned)
OPEN --[30 sec timeout]--> HALF_OPEN (testing)
HALF_OPEN --[1 success]--> CLOSED
HALF_OPEN --[1 failure]--> OPEN
Acceptance Criteria:
- New agents start with no circuit record (allowed)
- 5 failures within window trips circuit to OPEN
- OPEN agents receive HTTP 503 with
Retry-Afterheader - After 30 sec, OPEN transitions to HALF_OPEN
- Single success in HALF_OPEN closes circuit
- Single failure in HALF_OPEN re-trips to OPEN
- Admin can manually reset circuits
Known Limitations
- Performance Untested:
- EigenTrust computation on large graphs (>10k agents)
- Bloom filter memory at scale (currently sized for 1M items)
- Quarantine store scan performance with many pending items
- Circuit breaker store with many tripped agents
Commands to Run
# Full test suite
cargo test --workspace
# Phase 7 specific tests (176 total)
cargo test -p stemedb-storage -- circuit_breaker # 7D: 25 tests
cargo test -p stemedb-storage -- trust_graph # 7B: 23 tests
cargo test -p stemedb-storage -- trust_rank # Trust: 22 tests
cargo test -p stemedb-core -- pow # 7A: 19 tests
cargo test -p stemedb-storage -- domain_trust # 7B: 18 tests
cargo test -p stemedb-storage -- admission # 7A: 16 tests
cargo test -p stemedb-storage -- content_defense # 7C: 13 tests
cargo test -p stemedb-storage -- similarity_index # 7C: 12 tests
cargo test -p stemedb-storage -- quarantine # 7C: 9 tests
cargo test -p stemedb-core -- trust_tier # 7A: 8 tests
cargo test -p stemedb-api --test admission_integration # API: 6 tests
cargo test -p stemedb-ingest -- content_defense # 7C: 5 tests
# Clippy must pass
cargo clippy --workspace -- -D warnings
# Go SDK
cd sdk/go/steme && go test -v
Success Criteria
Phase 7 UAT passes when:
- All 176 Phase 7 tests pass
- All 7 usage scenarios verified manually
- Clippy clean with no warnings
- Go SDK tests pass
- API endpoints return correct responses
- Quarantine workflow complete end-to-end
- Circuit breaker state transitions verified
- Near-duplicate detection at 0.9 Jaccard threshold works