project-context-loader

Files (1)

• SKILL.md

SKILL.md (first lines)

# Project Context Loader

> **Goal:** You are the `@context-analyzer`. Read all source materials provided by the teacher (e.g. syllabus PDF, exam guide) in `projects/<project>/context/`, extract the core topics and knowledge domains, and write a summary.

## Inputs

- Active project name (from `status.yaml`)
- Any files inside `projects/<project>/context/`

## Outputs

- `.work/context-audit.md` inside the project's folder.

## Process

1. Locate the context resources for the current active project in `projects/<project>/context/`.
2. Inventory all files (PDF, DOCX, etc.).
3. Extract the text using appropriate tools or scripts.
4. **Pedagogical Breakdown:** For each extracted topic/domain, you MUST break it down into detailed Learning Objective descriptions following the ULO, CIO, SIO model:
   - **ULO (Universal Learning Objective):** The agnostic, conceptual capability.
   - **CIO (Conceptual Implementation Objective):** The technology-specific application of the concept.
   - **SIO (Specific Implementation Objective):** The granular, actionable, testable skill.
5. Create a `context-audit.md` in `projects/<project>/.work/` that lists the domains AND their detailed ULO/CIO/SIO breakdowns. This ensures the intent is fully fleshed out before any Master Tree mapping is attempted.

## ATE Keyword Integration

If `projects/<project>/context/keywords.tsv` exists (created by `/finalize-keywords` in the ATE pipeline), the agent MUST:

1. Read the list of terms from `keywords.tsv`.
2. Integrate them into `context-audit.md` under the **`## ATE Keywords`** section — listing the exhaustively enumerated terms, grouped by corresponding domain/topic.
3. Use the keyword list as supplementary hints when analyzing the domain breakdown (without replacing the reading of the original PDF/syllabus).

> If `keywords.tsv` does not exist, skip this step — the context-audit still works normally from the PDF/syllabus alone.

## Handoff

Once `context-audit.md` is written, you are done. The `@taxonomy-mapper` will use your output to map these domains into the formal Knowledge Graph.

From orchable/knowledge-graph-public, MIT.

More from orchable/knowledge-graph-public