wantsong
/
the-mindscape-of-bro-tsong
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
the-mindscape-of-bro-tsong/local_ccra_reviews/round-04/pilot-01/prompt-to-send.md

2.1 KiB
Raw Blame History

Prompt To Send: CCRA Local Review Pilot

You are being invoked as a real child review session for ccra-local-reviewer.

This is not main-session roleplay. If you cannot comply with the file-first invocation packet, report that you are blocked.

Read and follow:

C:\Users\wangq\Documents\Codex\work-projects\the-mindscape-of-bro-tsong\local_ccra_reviews\round-04\pilot-01\agent-invocation-packet.md

Canonical CCPE artifacts:

C:\Users\wangq\Documents\Codex\ccpe-system\agents\agent-specs\ccra-local-reviewer.agent.md
C:\Users\wangq\Documents\Codex\ccpe-system\runtimes\hybrid\ccra-local-review.runtime.md
C:\Users\wangq\Documents\Codex\ccpe-system\runtimes\hybrid\ccra-local-review-report.template.md

Required project knowledge base:

C:\Users\wangq\Documents\Codex\work-projects\the-mindscape-of-bro-tsong\knowledge_assets

Use this path directly. Do not create a second Local CCRA copy of the knowledge base.

Required output:

C:\Users\wangq\Documents\Codex\work-projects\the-mindscape-of-bro-tsong\local_ccra_reviews\round-04\pilot-01\04_LOCAL_CCRA_REVIEW_REPORT.md
C:\Users\wangq\Documents\Codex\work-projects\the-mindscape-of-bro-tsong\local_ccra_reviews\round-04\pilot-01\returned-output.md

Hard constraints:

  • Read local files directly.
  • Read the required knowledge_assets/ knowledge base.
  • Cite the files actually read.
  • Use helper outputs only as evidence, not as judgment.
  • Do not modify project files except the two required output files above.
  • Do not edit selector rules, model cards, model JSON, regression cases, review bundles, or Owner decision files.
  • Do not upgrade model lifecycle status.
  • Do not treat this pilot as Round 05 acceptance.

Pilot review questions:

  1. Was enough local evidence available to produce a stronger review than a limited Web upload?
  2. Does the Round 04 evidence show local tuning gaps that should be addressed before Round 05?
  3. What should be tuned in the Local CCRA Agent, Runtime, helper outputs, invocation packet, or report template before formal Round 05 use?
  4. Is this pilot report suitable for comparison against Web CCRA feedback supplied later by the Owner?