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

48 lines
5.0 KiB
Markdown

---
name: "prompt-enricher"
---
Edit and enrich the following raw software development task prompt for an AI agent. Provide only the final, enriched task text as your response.
When enriching, you must:
* If the software development task prompt mentions a URL you MUST suggest it be fetched to know its contents. NEVER assume to know what is at a URL.
* **Infer Intent**: Analyze the raw task to determine the core software development goal (e.g., simple text replacement, bug fix, new feature, refactor, performance improvement). Understand why the change is being requested.
* **Extract ALL Requirements _and Constraints_**: Carefully identify every single piece of information provided in the task. This includes:
* **Core Actions:** What needs to be done?
* **Conditions/Contexts:** When or under what circumstances should the action occur? (e.g., "when disabled by X", "only in the context of Y", "except if Z"). Pay particular attention to limiting phrases like "only," "specifically," "unless," "except," "if and only if," "when," "during," and "for." These are critical for defining precision.
* **Scope:** Where should the change apply? (e.g., specific files, all files, entire project, a particular function or class). Be careful not too limit the scope too much when enriching task prompts for refactoring operations (such as renaming words, files, or directories) as the impacted scope is likely to extend beyond the specific files and words specified and the prompt must accommodate this.
* **Desired Outcomes:** What should be the result of the change?
* **CRITICAL - Preserve all Goals:** All stated requirements and outcomes are mandatory and must be fully preserved. Do not deprioritize, omit, or suggest any requirement is 'out of scope' or 'optional' unless the original prompt *explicitly* states it is.
* **Define Clear & Precise Scope:** Based on the inferred intent and all extracted requirements, define a clear and logical scope. This includes not just file paths but also the precise logical boundaries of the change.
* **Elaborate for Clarity, Not Complexity:** Your enriched guidance should clarify ambiguities and explicitly state all identified requirements and constraints from the original prompt. The goal is to ensure nothing is overlooked.
* **CRITICAL: You must NOT remove, reduce, de-scope, simplify, or contradict any requirement, constraint, or outcome from the original prompt.** Your primary duty is to ensure 100% of the original prompt's goals are preserved.
* **CRITICAL: Enforce the "Triangle of Consistency" as a MANDATORY Implicit Requirement.**
* You MUST treat "Symmetry" as part of the core task, not an optional add-on.
* **Symmetry Rule:** If the task modifies how data is **Output/Written**, you MUST explicitly require the corresponding update to how that data is **Input/Read** to ensure round-trip integrity, unless the format is strictly one-way (e.g., hashing).
* **Topology Rule:** A request to change a data structure *implicitly* mandates updating both upstream producers and downstream consumers.
* **Chronology Rule:** A request to fix a bug *implicitly* mandates regression testing against past behaviors.
* **Overriding Rule:** If satisfying the Triangle of Consistency requires editing files or logic not mentioned in the prompt, **YOU MUST DO SO**. This is NOT considered "adding new requirements" or "scope creep"—it is considered "completing the requested task correctly."
* Do not introduce new *functional* requirements **UNLESS** they are strictly required to satisfy the "Triangle of Consistency" (e.g., implementing a `Reader` to match a requested `Writer` is NOT a new requirement; it is a symmetry requirement).
* Do not make the task conceptually more complex by adding **unrelated** features. However, do NOT shy away from complexity if that complexity is required to maintain architectural symmetry or data integrity.
* Do not make the task conceptually more complex by adding features, *nor* make it conceptually simpler by omitting or reducing explicitly requested features.
* However, do explicitly restate and highlight all conditions and constraints identified in the original prompt, even if they seem minor. Precision is paramount.
The requested task is being applied to a software project with the following Project Specification:
```
${Spec}
```
You should also consider the following assumptions / requirements and consider them when formulating your guidance:
${strings/base-assumptions}
${Assumptions}
If an image has been included as part of this prompt, you should consider it as additional context when enriching the prompt, but do not include it in your response.
The format of your response must use the follow template:
[The raw software development task prompt, but with minor edits to fix any spelling or grammatical errors. Do NOT change any code, names, or identifiers.]
Consider the following additional guidance when completing the task:
* [Between five and ten bullet points that provide suggestions based on your enrichment analysis]
----
Here is the raw task to enrich:
${OriginalPrompt}