Amanuensis

Evidence

Review coverage

See which failure modes were tested in each subsystem and the disposition reached for every applicable concern.

Branch  main Checked  adc4ce04aa57 As of  2026-08-29 04:56:43 No recorded stale entries

§Coverage overview

128Recorded reviews
8Subsystems
22Active concerns
8Confirmed defects

Only the 128 recorded reviews are expanded below; 48 unassessed intersections do not consume reading space. The thin rule shows recorded coverage; the labelled markers show review outcomes. 8 reviews are linchpin-dependent.

Confirmed defectAccepted behaviorRuled outOut of scopeCompeting explanationsLinchpin-dependentNot assessed
Full cross-reference matrix8 subsystems × 22 concerns

Use the matrix when exact row-and-column comparison matters. Headers and subsystem names remain fixed while the matrix scrolls.

Full concern review matrix
SubsystemAT-1AT-2CC-1CR-1EP-1EP-2IF-1RL-1RL-2SC-1SC-2SC-3SC-4SC-5SC-6SD-1SI-1SI-2TB-1TR-1TR-2VG-1
Survey methodology and agent contractsB-01Not assessedNot assessedNot assessedNot assessedNot assessed
MCP core, persistence, and lifecycleB-02Not assessedNot assessedNot assessedNot assessedNot assessedNot assessed
Knowledge tools and workflow APIB-03Not assessedNot assessedNot assessed
Diff-aware materializerB-04Not assessedNot assessedNot assessedNot assessed
Packaging, installer, validation, and product docsB-05Not assessedNot assessedNot assessedNot assessedNot assessed
Report interface design and validation studiesB-06Not assessedNot assessedNot assessed
Embedded research surveys and platform trialsB-07Not assessedNot assessedNot assessed
Activation evidence and release-readiness harnessB-08Not assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessedNot assessed

§Active concerns

AT-1

Category
atomicity
Origin
seeded
Discovered in

Notes

Inspect multi-table resets, status advances, finding transitions, and diagnosticity resolution for transaction boundaries covering all dependent writes.

AT-2

Category
atomicity
Origin
seeded
Discovered in

Notes

Verify a phase checkpoint persists SQLite WAL changes into the git-tracked memory.db before reporting a storage commit or no-op.

CC-1

Category
cache-coherence
Origin
seeded
Discovered in

Notes

Compare every materialized page with its declared DB/prose source set; omitted dependencies must not leave a page falsely unchanged.

CR-1

Category
concurrency
Origin
seeded
Discovered in

Notes

Verify concurrent agents, advisory write locks, SQLite WAL writers, and storage git commits cannot lose or misattribute landed work.

EP-1

Category
exceptional-paths
Origin
seeded
Discovered in

Notes

Verify a failed page render cannot publish a globally green or internally inconsistent documentation set.

EP-2

Category
exceptional-paths
Origin
seeded
Discovered in

Notes

Verify each mutating tool either commits all dependent rows or leaves pre-call state intact when validation, SQL, or filesystem work fails.

IF-1

Category
incremental-full-divergence
Origin
seeded
Discovered in

Notes

Compare incremental materialization, force-full materialization, and clean-export outputs for exact state, coverage, and content correspondence.

RL-1

Category
resource-lifecycle
Origin
seeded
Discovered in

Notes

Verify database handles, spawned processes, temporary directories, and materializer locks release on success, error, timeout, and cancellation.

RL-2

Category
resource-lifecycle
Origin
seeded
Discovered in

Notes

Verify sessions, dispatches, open questions, and write locks have explicit terminal or recovery paths and cannot remain silently authoritative after abandonment.

SC-1

Category
seam-contract
Origin
seeded
Discovered in

Notes

Verify schema columns/views, TypeScript tool queries, and Python materializer reads evolve together and fail closed on contract drift.

SC-2

Category
seam-contract
Origin
seeded
Discovered in

Notes

Verify package-time agent/materializer mirrors are derived from root sources and CI detects stale or host-dependent packaged assets.

SC-3

Category
seam-contract
Origin
discovered
Discovered in
SM-06

Notes

Verify the B-06 report-presentation contract and B-04 typed HTML projection agree on record-vs-table selection, label translation, identifier definitions, and revision-bound read-back.

SC-4

Category
seam-contract
Origin
discovered
Discovered in
SM-07

Notes

Verify the B-06 product ruling/design language and B-01 reporting-style contract agree on project identity, practitioner register, specialized labels, record grammar, and identifier treatment.

SC-5

Category
seam-contract
Origin
discovered
Discovered in
SM-08

Notes

Verify B-07 research limitations survive transfer into B-06 report-design rules: report-owned re-entry must not imply personal history, and corpus/prototype evidence must not be promoted to reader-performance or conformance guarantees.

SC-6

Category
seam-contract
Origin
discovered
Discovered in
B-03

Notes

Verify the evidence-quality vocabulary the B-01 methodology contract publishes is the vocabulary the B-03 tool layer accepts, so that recording the strongest kind an agent's evidence actually supports is always expressible rather than forcing an over- or under-statement.

SD-1

Category
scope-coverage
Origin
discovered
Discovered in
B-03

Notes

Verify the file ledger is reconciled against the working tree: files added since the last baseline must be surfaced as unclassified, and ledger rows for deleted files must not remain authoritative as 'examined'. A subsystem reported 'mapped' with 0 stale entries must not be able to conceal unclassified or nonexistent scope.

SI-1

Category
scope-identity
Origin
seeded
Discovered in

Notes

Verify project-key derivation, stored workspace_path collision detection, and storage-root confinement cannot conflate two workspaces or escape the root.

SI-2

Category
revision-identity
Origin
seeded
Discovered in

Notes

Verify every evidence-bearing record and generated artifact retains the exact repository revision and cannot be reused as current evidence at another SHA.

TB-1

Category
temporal-bounds
Origin
seeded
Discovered in

Notes

Inspect every git and Python subprocess invocation for bounded completion, cancellation, and actionable failure propagation.

TR-1

Category
trust-boundary
Origin
seeded
Discovered in

Notes

Verify MCP string/path/enum inputs cannot escape storage, inject SQL or git arguments, or create authority beyond the active subsystem depth.

TR-2

Category
trust-boundary
Origin
seeded
Discovered in

Notes

Verify installer parsing and config merging preserve unrelated user configuration, back up destructive overwrites, and reject path traversal.

VG-1

Category
validation-gate
Origin
discovered
Discovered in
B-03

Notes

Verify every published health signal is computed from a source that production code actually populates. A gate whose denominator is structurally always zero reports green regardless of system state and is not evidence of health.