Evidence ledger
この ledger は、公開 repository から再現できる evidence を列挙します。private operator logs、real chat transcripts、secret-bearing configs、raw internal review notes は含めません。
Core gates
Section titled “Core gates”| Evidence | Command | Supports |
|---|---|---|
| Formatting invariant | cargo fmt -- --check |
Release hygiene |
| Rust warnings-as-errors | cargo clippy -- -D warnings |
Baseline code quality |
| Full default test matrix | cargo test |
Broad regression coverage |
| Project policy tests | cargo test --test project |
Architecture, release policy, fixture inventory |
| Docs build | pnpm --dir docs build |
Public documentation coherence |
| Cargo metadata | cargo metadata --no-deps --format-version 1 |
Package license / repository / readme metadata |
Claim-focused checks
Section titled “Claim-focused checks”| Claim area | Command |
|---|---|
| Transport turn contract | cargo test --test runtime companion_turn_contract |
| Architecture boundaries | cargo test --test project module_boundaries |
| Memory behavior | cargo test --test memory |
| Runtime security guarantees | cargo test --test runtime security_guarantees |
| Gateway auth / scope | cargo test --test gateway auth |
| Naturalness / pre-send fixtures | cargo test --lib naturalness |
| Companion harness OFF / ON ablation | cargo run -- eval harness --fixtures tests/fixtures/harness --output evidence/harness-ablation-report.json |
| Model-backed companion harness ablation | cargo run -- eval harness --fixtures tests/fixtures/harness --model-backed --provider <provider> --model <model> --temperature 0.4 --output evidence/harness-model-backed-report.json |
| Bad-turn replay fixture | cargo run -- eval replay --input tests/fixtures/replay/discord_companion_bad_turns.jsonl --suite discord-companion-bad-turns |
| Architecture helper scripts | ./scripts/dev/generate_module_map.sh && ./scripts/dev/check_architecture.sh |
Publishable result forms
Section titled “Publishable result forms”Evidence として公開してよいもの:
- command name、commit hash、toolchain version、pass / fail result
- aggregate counts、fixture IDs、verifier reason names、failure taxonomy
- real user / private memory を含まない synthetic fixture text
- 再配布可能な synthetic または licensed benchmark prompts / scoring rubrics
- tokens、tenant / person IDs、raw memory payloads、provider responses を除いた redacted logs
公開しないもの:
- private Discord logs や real relationship memory
- private context を含む provider prompts / responses
- OAuth / client secrets、provider keys、pairing tokens、webhook signatures
- unresolved security review findings や exploit details
- operator environment を特定する local machine paths
- internal handoff prompts や raw design-debate notes