SikongJueluo 72c9366f07 test(ai-judge): deepseek v4 flash catalog qualification ...

Two strict corpus-replay rounds against deepseek/deepseek-v4-flash
(api openai-completions, 30000ms timeout):

- run -01: 20/21, covered-compound judged defer (non-repeating
  sampling miss, not systematic)
- run -02: 21/21, zero infrastructure failures, p50 1448ms /
  p95 1682ms / max 2146ms

Qualified. Adds the catalog entry (recommended, advisory data per
ADR 0008) and retains both reports; run -01 notes the single miss.
Full repo check+test green (236 + 48).

2026-08-21 23:41:26 +08:00

.pi

docs(ai-judge): add package README

2026-08-21 23:11:34 +08:00

docs

feat(ai-judge): rework enforce mode as user-assumed-risk contract

2026-08-21 23:11:40 +08:00

packages

test(ai-judge): deepseek v4 flash catalog qualification

2026-08-21 23:41:26 +08:00

.gitignore

docs(testing): agent-driven TUI replay flow with 30-command cohort run

2026-08-18 00:33:47 +08:00

AGENTS.md

docs: add agent guidance and permission research docs

2026-08-09 00:23:24 +08:00

CONTEXT.md

feat(ai-judge): rework enforce mode as user-assumed-risk contract

2026-08-21 23:11:40 +08:00

LICENSE

Initial commit

2026-08-08 18:14:37 +08:00

package.json

build: align permission-system to 25.4

2026-08-17 19:20:49 +08:00

pnpm-lock.yaml

build: align permission-system to 25.4

2026-08-17 19:20:49 +08:00

pnpm-workspace.yaml

refactor(permission): consume structured bash payload

2026-08-16 23:08:45 +08:00

tsconfig.base.json

feat(pi-permission-ai-judge): add deferring authorizer extension

2026-08-08 19:43:21 +08:00

S

Description

No description provided

GPL-3.0

496 KiB

Languages

TypeScript 97.7%

JavaScript 2.3%