Claude 24f816b6a3
Consolidate 22 sibling repos into layered organism structure
Place useful parts of the surrounding repos into sica-fondt by layer, per the
body model (Ada = membrane; brain/endocrine/capabilities/knowledge non-Ada):

- brain/        LLM reasoning + providers (dapr, hermes, MoMoA)
- capabilities/ REPRAG sidecars: hermes tools/skills, dapr tools, parallel
                dispatch, A51 channels, and the OSINT cluster
- knowledge/    LORAG corpus: 754 cyber-skills, agency personas, secure-coding,
                MITRE ATT&CK data
- reference/    defensive threat-reference (C3, shhbruh doc) + AdaYaml parser

License handling: AGPL sources (worldosint, advanced_evolution, mercury,
Reticulum) and GPL DeTTECT are SPEC-only clean-room/port descriptions — no
copyleft code copied. MIT/Apache/data parts copied as working trees.

Safety: shhbruh escape/persistence material and C3 covert-C2 kept as reference
only, not wired into the running organism. See CONSOLIDATION.md.

https://claude.ai/code/session_01UehUqEXXJJCsHoA4voCU5c
2026-06-10 06:53:01 +00:00

25 lines
1.5 KiB
Markdown

---
name: "summaryXMLPrompt"
---
Generate a final project summary in valid XML format based on the provided XML summary, and the following specific instructions.
The XML must have a root element named '<Summary>'.
It must contain the following child elements:
- <RESULT>: A summary of the project's outcome or main deliverable.
- <WORK_SUMMARY>: A retrospective summary of the work performed by the experts.
- <USER_FEEDBACK>: Feedback intended for the user about the project outcome.
- <DEV_FEEDBACK>: Feedback intended for the agent/tool developers about the process.
- <FILES>: A list of files that were produced or significantly modified. Each file should be represented by a <FILE> element containing a <FILE_NAME> element with the filename as its text content and a <DESCRIPTION> element with a brief description of the file's contents.
- <MODEL_USE>: A summary of how many times each LLM API/Model pair was invoked during the process.
Ensure the content within the <RESULT>, <WORK_SUMMARY>, <USER_FEEDBACK>, and <DEV_FEEDBACK> tags is plain text (avoiding internal XML tags unless properly escaped or in CDATA).
The <FILES> section should list all the files that were produced, based on the available files.
Here is the required structure for the <MODEL_USE> element. Include this structure as a child of the <Summary> element, with appropriate line breaks and indentation to make it easy to read:
${messageCountsXml}
The final response from the orchestrator provided below:
${finalOrchestratorResult}
Generate the complete XML summary now, starting with the <Summary> tag.