repr-cli 0.2.15__py3-none-any.whl → 0.2.17__py3-none-any.whl

repr/templates.py

@@ -16,12 +16,27 @@ class StoryOutput(BaseModel):
     """Structured output for a generated story."""
     summary: str = Field(description="One-line technical summary of the work (max 120 chars, no fluff)")
     content: str = Field(description="Full technical description in markdown")
+    # Technologies used in this work
+    technologies: list[str] | None = Field(default=None, description="Languages, frameworks, libraries, tools used (e.g., TypeScript, React, PostgreSQL, Docker)")
     # Categories - all inferred by LLM, nullable for backwards compatibility with older stories
     category: str | None = Field(default=None, description="Work type: feature, bugfix, refactor, perf, infra, docs, test, chore")
     scope: str | None = Field(default=None, description="Impact scope: user-facing, internal, platform, ops")
     stack: str | None = Field(default=None, description="Stack layer: frontend, backend, database, infra, mobile, fullstack")
     initiative: str | None = Field(default=None, description="Initiative type: greenfield, migration, integration, scaling, incident-response, tech-debt")
     complexity: str | None = Field(default=None, description="Complexity: quick-win, project, epic, architecture")
+    
+    # Context for AI agents - captures the WHY/HOW/DECISIONS
+    problem: str | None = Field(default=None, description="What problem was being solved?")
+    approach: str | None = Field(default=None, description="What pattern/approach was used?")
+    decisions: list[str] | None = Field(default=None, description="Key decisions made (format: 'Chose X over Y because Z')")
+    tradeoffs: str | None = Field(default=None, description="What was gained/lost?")
+    outcome: str | None = Field(default=None, description="Metrics, before/after, observable impact")
+    lessons: list[str] | None = Field(default=None, description="Gotchas, learnings, things to remember")
+    
+    # Source tracking for timeline entries
+    source_type: str | None = Field(default=None, description="'commit', 'session', or 'merged'")
+    session_id: str | None = Field(default=None, description="Linked session ID if from AI session")
+    commit_sha: str | None = Field(default=None, description="Linked commit SHA if from git")
 
 
 # Template definitions
@@ -36,6 +51,9 @@ Output JSON with:
   Good: "Added JWT refresh token rotation with Redis session store"
   Bad: "Enhanced authentication system with improved security"
 - content: Markdown with technical details. What was built, how, what tech.
+- technologies: Array of languages, frameworks, libraries, tools used.
+  Examples: ["TypeScript", "React", "PostgreSQL", "Docker", "Redis", "GraphQL"]
+  Be specific: "React" not "JavaScript framework", "FastAPI" not "Python web framework"
 
 Rules:
 - Name specific technologies, libraries, patterns
@@ -48,7 +66,7 @@ Rules:
 Commits:
 {commits_summary}
 
-Output JSON with summary and content.""",
+Output JSON with summary, content, and technologies.""",
     },
     
     "changelog": {
@@ -59,6 +77,7 @@ Output JSON with summary and content.""",
 Output JSON with:
 - summary: One line describing the main change (max 120 chars)
 - content: Markdown changelog with categories (Added/Changed/Fixed/Removed)
+- technologies: Array of languages, frameworks, libraries, tools involved in these changes
 
 Rules:
 - List actual changes, not benefits
@@ -69,7 +88,7 @@ Rules:
 Commits:
 {commits_summary}
 
-Output JSON with summary and content.""",
+Output JSON with summary, content, and technologies.""",
     },
     
     "narrative": {
@@ -80,6 +99,7 @@ Output JSON with summary and content.""",
 Output JSON with:
 - summary: One-line description of what was built (max 120 chars)
 - content: Markdown narrative explaining the technical work
+- technologies: Array of specific languages, frameworks, libraries, tools used
 
 Focus on:
 - What problem was solved
@@ -92,7 +112,7 @@ No marketing language. Write like you're explaining to another engineer.""",
 Commits:
 {commits_summary}
 
-Output JSON with summary and content.""",
+Output JSON with summary, content, and technologies.""",
     },
     
     "interview": {
@@ -103,6 +123,7 @@ Output JSON with summary and content.""",
 Output JSON with:
 - summary: One-line technical summary (max 120 chars)
 - content: Markdown with situation/task/action/result format
+- technologies: Array of specific languages, frameworks, libraries, tools used
 
 Focus on:
 - Specific technical decisions made
@@ -115,7 +136,50 @@ No resume language. Be specific about what you actually did.""",
 Commits:
 {commits_summary}
 
-Output JSON with summary and content.""",
+Output JSON with summary, content, and technologies.""",
+    },
+    
+    "context": {
+        "name": "Context",
+        "description": "Rich context for AI agents - captures WHY, HOW, and DECISIONS",
+        "system_prompt": """Extract developer context from commits and sessions.
+
+Your output will be consumed by AI coding assistants that need to understand this developer's patterns, decisions, and reasoning.
+
+For each piece of work, capture:
+
+1. WHAT: One-line summary of what was built/changed
+2. WHY (problem): What problem was this solving? What was broken/slow/missing?
+3. HOW (approach): What strategy/pattern was used? What does the code do?
+4. TECHNOLOGIES: What specific languages, frameworks, libraries, tools were used?
+5. DECISIONS: Were there alternatives? Why this approach? Format each as "Chose X over Y because Z"
+6. TRADEOFFS: What did you give up? What did you gain?
+7. OUTCOME: Any metrics, before/after, or observable impact?
+8. LESSONS: What gotchas were discovered? What would you tell future-you?
+
+Write for an AI that will answer questions like:
+- "How does this developer usually handle X?"
+- "What patterns do they use for Y?"
+- "Why did they choose Z over alternatives?"
+
+Output valid JSON with all fields from the schema.""",
+        "user_prompt_template": """Repository: {repo_name}
+
+Commits:
+{commits_summary}
+
+Extract structured context with:
+- summary: One-line technical summary (max 120 chars)
+- content: Full markdown description
+- technologies: Array of languages, frameworks, libraries, tools (e.g., ["TypeScript", "React", "PostgreSQL"])
+- problem: What problem was being solved?
+- approach: What pattern/strategy was used?
+- decisions: List of key decisions (format: "Chose X over Y because Z")
+- tradeoffs: What was gained/lost?
+- outcome: Results, metrics if available
+- lessons: Gotchas and learnings
+
+Output valid JSON only.""",
     },
 }