feat: add sanitized lighthouse skill harbor agent
This commit is contained in:
parent
a68945789b
commit
e43e54b195
20 changed files with 829 additions and 5 deletions
10
server-tools/fifth-domain-lighthouse-mirror/README.md
Normal file
10
server-tools/fifth-domain-lighthouse-mirror/README.md
Normal file
|
|
@ -0,0 +1,10 @@
|
|||
# Fifth Domain Lighthouse Sanitized Mirror Agent
|
||||
|
||||
This deterministic Agent has two independent functions:
|
||||
|
||||
1. Build a field-allowlisted, sanitized local mirror of registered public language architecture from the committed REPO-012 snapshot.
|
||||
2. Accept explicitly shared persona skill contributions into quarantine, bind mother and Guanghu human-team reviews to the exact candidate digest, and register only double-accepted candidates.
|
||||
|
||||
It never copies a private skill pack, never executes contributed code, never publishes externally, and grants no authority. The shared runtime defaults to `persona-runtime/shared/lighthouse-persona-skill-harbor`; tests must use a temporary `--state-root`.
|
||||
|
||||
Commands: `sync-architecture`, `prepare`, `mother-review`, `team-review`, `register`, and `status`.
|
||||
Loading…
Reference in a new issue