Rockachopa/Timmy-time-dashboard
1
0
Fork 2
Code Issues 7 Pull Requests Actions Packages Projects Releases Wiki Activity

[P0] Build ResearchOrchestrator pipeline (src/timmy/research.py) #975

New Issue
Closed
opened 2026-03-22 19:08:53 +00:00 by perplexity · 1 comment
perplexity commented 2026-03-22 19:08:53 +00:00
Collaborator
Copy Link

Parent

  • #972 — [GOVERNING] Replacing Claude — Autonomous Research Pipeline Spec

Objective

Implement the main research pipeline that chains Scope→Query→Search→Fetch→Synthesize→Deliver into an autonomous workflow. This is the core of research sovereignty.

Scope

Implement ResearchOrchestrator class with:

class ResearchOrchestrator:
    def __init__(self, cascade, memory, tools)
    async def run(self, topic, template, context) -> ResearchResult

Pipeline Steps:

  1. CHECK LOCAL KNOWLEDGE FIRST — memory.search(topic, limit=10). If confidence > 0.85, return cached result. This is the critical line.
  2. GENERATE QUERIES — Fill template, ask cascade to generate 8-12 search queries
  3. SEARCH — Execute queries via web_search, collect top 5 results per query
  4. FETCH — Rank snippets by relevance, fetch top 10 full pages via web_fetch (3000 tokens each)
  5. SYNTHESIZE — Pass filled template + fetched pages to cascade.generate (max_tokens=4000)
  6. CRYSTALLIZE — Store result in semantic memory (memory.store(topic, report, type="research"))
  7. WRITE ARTIFACT — Commit to repo, extract action items, create Gitea issues

Dependencies

  • web_fetch tool (sibling P0 issue)
  • Research templates (sibling P0 issue)
  • cascade.py (exists — LLM router)
  • Semantic memory (exists or P1 enhancement)

Key Design Notes

  • Runs as Paperclip task via DistributedWorker (P1 integration)
  • The cascade router passes template's cascade_tier hint for model selection
  • Every research output gets embedded → compound interest of crystallized knowledge
  • Record metrics: research_cache_hit, research_api_call

Effort Estimate

1 day

Acceptance Criteria

  • Pipeline runs end-to-end: topic in → structured report out
  • Local knowledge check works (cache hit returns instantly)
  • Results are stored in semantic memory for future queries
  • Gitea issues are created from extracted action items
  • Metrics are recorded for sovereignty tracking
## Parent - #972 — [GOVERNING] Replacing Claude — Autonomous Research Pipeline Spec ## Objective Implement the main research pipeline that chains Scope→Query→Search→Fetch→Synthesize→Deliver into an autonomous workflow. This is the core of research sovereignty. ## Scope Implement `ResearchOrchestrator` class with: ``` class ResearchOrchestrator: def __init__(self, cascade, memory, tools) async def run(self, topic, template, context) -> ResearchResult ``` ### Pipeline Steps: 0. **CHECK LOCAL KNOWLEDGE FIRST** — `memory.search(topic, limit=10)`. If confidence > 0.85, return cached result. This is the critical line. 1. **GENERATE QUERIES** — Fill template, ask cascade to generate 8-12 search queries 2. **SEARCH** — Execute queries via web_search, collect top 5 results per query 3. **FETCH** — Rank snippets by relevance, fetch top 10 full pages via web_fetch (3000 tokens each) 4. **SYNTHESIZE** — Pass filled template + fetched pages to cascade.generate (max_tokens=4000) 5. **CRYSTALLIZE** — Store result in semantic memory (`memory.store(topic, report, type="research")`) 6. **WRITE ARTIFACT** — Commit to repo, extract action items, create Gitea issues ## Dependencies - web_fetch tool (sibling P0 issue) - Research templates (sibling P0 issue) - cascade.py (exists — LLM router) - Semantic memory (exists or P1 enhancement) ## Key Design Notes - Runs as Paperclip task via DistributedWorker (P1 integration) - The cascade router passes template's `cascade_tier` hint for model selection - Every research output gets embedded → compound interest of crystallized knowledge - Record metrics: `research_cache_hit`, `research_api_call` ## Effort Estimate 1 day ## Acceptance Criteria - [ ] Pipeline runs end-to-end: topic in → structured report out - [ ] Local knowledge check works (cache hit returns instantly) - [ ] Results are stored in semantic memory for future queries - [ ] Gitea issues are created from extracted action items - [ ] Metrics are recorded for sovereignty tracking
claude was assigned by Rockachopa 2026-03-22 21:44:41 +00:00
Rockachopa commented 2026-03-22 22:44:48 +00:00
Owner
Copy Link

PR #1000 created.

Implemented ResearchOrchestrator in src/timmy/research.py with full 7-step pipeline: local knowledge check → query generation → web search → fetch → synthesize → crystallize → write artifact. Includes 25 unit tests, graceful degradation at every step, and Gitea issue creation from extracted action items.

PR #1000 created. Implemented `ResearchOrchestrator` in `src/timmy/research.py` with full 7-step pipeline: local knowledge check → query generation → web search → fetch → synthesize → crystallize → write artifact. Includes 25 unit tests, graceful degradation at every step, and Gitea issue creation from extracted action items.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
222-epic
Workshop: Timmy as Presence (Epic #222)
 actionable
Has a concrete code/config task extracted
 assigned-claude
Issue currently assigned to Claude agent — do not assign to another agent
 assigned-gemini
Issue currently assigned to Gemini agent — do not assign to another agent
 assigned-groq
assigned-kimi
Issue currently assigned to Kimi agent — do not assign to another agent
 assigned-manus
Issue currently assigned to Manus agent — do not assign to another agent
 claude-ready
consolidation
Part of a consolidation epic
 deprioritized
Keep open but not blocking P0 work
 deprioritized
Keep open but not blocking P0 work
 duplicate
Duplicate of another issue
 gemini-review
Auto-generated by Gemini, needs relevance review
 groq-ready
harness
Core product: agent framework, heartbeat, inference, memory
 heartbeat
Harness: Agent heartbeat loop
 inference
Harness: Inference and model routing
 infrastructure
Supporting stage: dashboard, CI/CD, deployment, DNS
 kimi-ready
Scoped and ready for Kimi to pick up
 memory-session
Harness: Memory and session crystallization
 morrowind
Harness: Morrowind embodiment
 needs-design
Needs architectural design before implementation
 needs-extraction
Philosophy with unextracted engineering work
 p0-critical
Priority 0: Must fix now
 p1-important
Priority 1: Important, next sprint
 p2-backlog
Priority 2: Backlog, do when time permits
 philosophy
Philosophical foundation — informs architecture decisions
 rejected-direction
Closed: rejected or superseded direction
 seed:know-purpose
Three Seeds: KNOW YOUR PURPOSE
 seed:serve-real
Three Seeds: SERVE THE REAL
 seed:tell-truth
Three Seeds: TELL THE TRUTH
 sovereignty
Harness: Sovereignty stack
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
Timmy claude grok groq kimi manus perplexity Rockachopa
No Assignees claude
2 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Rockachopa/Timmy-time-dashboard#975
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?