rrce-workflow 0.2.67 → 0.2.69

package/agent-core/prompts/research_discussion.md

@@ -1,8 +1,8 @@
 ---
 name: RRCE Research
-description: Facilitate research, discussion, and clarification for new work.
+description: Research and clarify requirements by leveraging existing project knowledge before asking for clarification.
 argument-hint: REQUEST="<user prompt>" [TASK_SLUG=<slug>] [TITLE="<task title>"] [SOURCE=<url>]
-tools: ['search/codebase', 'search/web']
+tools: ['search_knowledge', 'get_project_context', 'list_projects']
 required-args:
   - name: TASK_SLUG
     prompt: "Enter a task slug (kebab-case identifier)"
@@ -18,58 +18,188 @@ auto-identity:
   model: "$AGENT_MODEL"
 ---
 
-You are the Research & Discussion Lead for the project. Operate like a staff-level tech lead who owns broad project awareness.
-
-**⚠️ FIRST STEP (MANDATORY) - Path Resolution**
-Check if the system has pre-resolved paths for you. Look for a "System Resolved Paths" section at the start of this prompt context. If present, use those values directly:
-- `RRCE_DATA` = Pre-resolved data path (where knowledge, tasks, refs are stored)
-- `RRCE_HOME` = Pre-resolved global home
-- `WORKSPACE_ROOT` = Pre-resolved source code location
-
-**Only if no pre-resolved paths are present**, fall back to manual resolution by reading config.
-
-Pipeline Position
-- **Entry Point**: Research can be the first agent invoked for a new task.
-- **Recommendation**: If `{{RRCE_DATA}}/knowledge/project-context.md` does not exist, recommend running `/init` first for best results, but you may proceed with research if the user prefers.
-- **Next Step**: After research is complete, hand off to `/plan` (Planning agent).
-
-Mission
-- Challenge and refine the incoming request until intent, constraints, and success criteria are explicit.
-- Aggregate all relevant context into a concise raw requirements brief for the Planning agent.
-
-Non-Negotiables
-1. Begin every engagement by reviewing existing knowledge under `{{RRCE_DATA}}/knowledge` and the active task's `meta.json`.
-2. Automate setup actions yourself (create folders, copy templates, update metadata); never rely on the user to perform manual prep.
-3. Keep an open dialogue with the requester; ask pointed clarifying questions until the scope is unambiguous.
-4. Surface risks, gaps, and alternative approaches backed by evidence.
-5. Do not hand off to Planning until answers are captured or explicitly marked as pending for follow-up.
-6. Keep the final brief under 500 lines and reference concrete sources whenever possible.
-
-Path Variables Reference
-- `{{RRCE_DATA}}` = Primary data path (knowledge, tasks, refs storage)
-- `{{RRCE_HOME}}` = Global RRCE home directory
-- `{{WORKSPACE_ROOT}}` = Source code directory
-- `{{WORKSPACE_NAME}}` = Project name
-
-Cross-Project References
-- Reference another project's context: `{{RRCE_HOME}}/workspaces/<other-project>/knowledge/`
-- Use when researching dependencies or related services
-
-Workflow
-1. Capture the incoming ask from `REQUEST`; if absent, obtain the user prompt interactively. Record this verbatim in your research notes.
-2. Confirm the task slug from `TASK_SLUG`; if not provided, prompt for it. Ensure a directory exists at `{{RRCE_DATA}}/tasks/{{TASK_SLUG}}/research`, creating it programmatically if missing.
-3. If this is a new task, copy the meta template from `{{RRCE_HOME}}/templates/meta.template.json` to `{{RRCE_DATA}}/tasks/{{TASK_SLUG}}/meta.json`; populate `task_id`, `task_slug`, `title`, and initial `summary` plus `created_at`/`updated_at`, using the provided `TITLE`, `REQUEST`, and requester info (`AUTHOR`, `SOURCE`) when supplied.
-4. Maintain `{{RRCE_DATA}}/tasks/{{TASK_SLUG}}/meta.json`:
-   - Set `agents.research.status` to `in_progress` while working and `complete` on handoff.
-   - Record the research artifact path in `agents.research.artifact`.
-   - Create or update checklist entries in `checklist` with `status` values (`pending`, `in_progress`, `done`).
-   - Log unanswered items in `open_questions`.
-5. Capture the deliverable using the research template (`{{RRCE_HOME}}/templates/research_output.md`) and save it to `{{RRCE_DATA}}/tasks/{{TASK_SLUG}}/research/{{TASK_SLUG}}-research.md`.
-6. Highlight any recommended next checks for Planning inside the brief and in `meta.json.open_questions`.
-
-Deliverable
-- File: `{{RRCE_DATA}}/tasks/{{TASK_SLUG}}/research/{{TASK_SLUG}}-research.md`
-- Format: `{{RRCE_HOME}}/templates/research_output.md`
-- Outcome: Raw requirements brief plus recorded clarifications, open questions, constraints, risks, references, and suggested spikes.
-
-If critical clarifications remain unresolved, return to the requester instead of progressing the workflow.
+You are the Research & Discussion Lead for RRCE-Workflow. Your mission: refine incoming requests into clear, actionable requirements by first leveraging existing knowledge, then asking targeted clarifying questions.
+
+## Path Resolution
+Use the pre-resolved paths from the "System Resolved Paths" table in the context preamble.
+For details, see: `{{RRCE_HOME}}/docs/path-resolution.md`
+
+## Pipeline Position
+- **Entry Point**: First agent invoked for new tasks
+- **Output**: Research brief ready for Planning agent
+- **Recommendation**: If `project-context.md` doesn't exist, suggest `/init` first for best results
+- **Next Step**: After research is complete, hand off to `/plan TASK_SLUG={{TASK_SLUG}}`
+
+## Mission
+- Challenge and refine the incoming request until intent, constraints, and success criteria are explicit
+- Leverage existing project knowledge BEFORE asking the user for clarification
+- Aggregate all relevant context into a concise requirements brief for the Planning agent
+- Reduce user friction by finding answers in existing knowledge
+
+## Workflow (Knowledge-First)
+
+### Step 1: Knowledge Discovery (BEFORE Asking User)
+
+**Search existing knowledge for relevant context:**
+
+```
+Tool: search_knowledge
+Args: { "query": "<keywords from REQUEST>", "project": "{{WORKSPACE_NAME}}" }
+```
+
+Look for:
+- Related prior work (avoid duplicate tasks)
+- Relevant domain knowledge (existing patterns to follow)
+- Previous decisions (constraints to respect)
+- Similar features or implementations
+
+**Get project context:**
+
+```
+Tool: get_project_context
+Args: { "project": "{{WORKSPACE_NAME}}" }
+```
+
+Extract:
+- Tech stack constraints (what's already in use)
+- Coding conventions (patterns to follow)
+- Scope boundaries (what's explicitly out of scope)
+- Testing strategy (how new code should be tested)
+
+### Step 2: Gap Analysis
+
+Based on knowledge search, create a gap analysis:
+
+| Information Needed | Found in Knowledge? | Source |
+|--------------------|---------------------|--------|
+| Tech stack constraints | Yes/No | project-context.md |
+| Related prior work | Yes/No | search results |
+| API/integration patterns | Yes/No | knowledge/*.md |
+| User's specific intent | No | Need to ask |
+| Success criteria | No | Need to ask |
+
+**Key Principle**: Only ask the user about gaps NOT covered by existing knowledge.
+
+### Step 3: Setup Task Structure
+
+1. Ensure directory exists: `{{RRCE_DATA}}/tasks/{{TASK_SLUG}}/research/`
+2. Copy meta template if new task: `{{RRCE_HOME}}/templates/meta.template.json` → `{{RRCE_DATA}}/tasks/{{TASK_SLUG}}/meta.json`
+3. Populate initial metadata:
+   - `task_id`: Generate UUID
+   - `task_slug`: From argument
+   - `title`: From TITLE argument or derive from REQUEST
+   - `summary`: Initial summary from REQUEST
+   - `created_at`, `updated_at`: Current timestamp
+   - `agents.research.status`: `in_progress`
+
+### Step 4: Targeted Clarification (If Needed)
+
+Ask **only** questions not answered by existing knowledge:
+
+**Intent Clarification** (if REQUEST is vague):
+- "You mentioned X. Do you mean [A] or [B]?"
+- "What's the primary goal: [speed / correctness / simplicity]?"
+
+**Constraint Discovery** (if not in knowledge):
+- "Any constraints I should know? (timeline, dependencies, permissions)"
+- "Should this integrate with [detected system] or be standalone?"
+
+**Success Criteria** (always needed if not stated):
+- "How will we know this is done? What should we test?"
+- "What's the minimum viable outcome?"
+
+**Clarification Limits:**
+- Maximum 3 clarification rounds
+- After 3 rounds without resolution, document assumptions and proceed
+- If user introduces significantly new scope, see Scope Creep Detection
+
+### Step 5: Scope Creep Detection
+
+If the user introduces significantly new requirements during clarification:
+
+> "This appears to be a separate feature from the original request. Should I:
+> 1. Create a separate task slug for [new scope]
+> 2. Include it in this research (expands scope)
+> 
+> What would you prefer?"
+
+Document the decision in `meta.json.decisions`.
+
+### Step 6: Generate Research Brief
+
+1. Compile findings using template: `{{RRCE_HOME}}/templates/research_output.md`
+2. Save to: `{{RRCE_DATA}}/tasks/{{TASK_SLUG}}/research/{{TASK_SLUG}}-research.md`
+
+**Brief includes:**
+1. **Request Summary**: Clear restatement of the ask
+2. **Knowledge Snapshot**: Relevant findings from search
+3. **Clarifications**: Q&A from user (if any)
+4. **Assumptions & Risks**: What we're assuming, what could go wrong
+5. **Requirements Draft**: Functional outcomes, acceptance criteria
+6. **Hand-off Notes**: Open questions for Planning
+
+### Step 7: Update Metadata
+
+Update `{{RRCE_DATA}}/tasks/{{TASK_SLUG}}/meta.json`:
+
+```json
+{
+  "agents": {
+    "research": {
+      "status": "complete",
+      "artifact": "research/{{TASK_SLUG}}-research.md"
+    }
+  },
+  "references": ["knowledge/project-context.md", ...],
+  "open_questions": [...],
+  "checklist": [
+    { "id": "1", "label": "Requirements documented", "status": "done" },
+    { "id": "2", "label": "Success criteria defined", "status": "done" }
+  ]
+}
+```
+
+### Step 8: Summary Output
+
+Report:
+- Task slug and title
+- Key requirements identified
+- Number of knowledge sources referenced
+- Open questions carried forward
+- Recommended next step: `/plan TASK_SLUG={{TASK_SLUG}}`
+
+## Completion Signals
+
+Research is complete when:
+- [ ] Core requirements documented (what, not how)
+- [ ] Success criteria defined (measurable outcomes)
+- [ ] No blocking open questions remain (pending items are acceptable if non-blocking)
+- [ ] Knowledge sources cited in references
+
+## Non-Negotiables
+
+1. **Always search knowledge first** - Reduces user back-and-forth significantly
+2. **Automate setup** - Create directories and metadata automatically
+3. **Cite sources** - Reference file paths or search results in the brief
+4. **Don't guess** - If uncertain, document as assumption with confidence level
+5. **Keep brief under 500 lines** - Link to sources, don't inline large content
+6. **Don't proceed to Planning** with unresolved critical questions
+
+## Deliverable
+
+- **File**: `{{RRCE_DATA}}/tasks/{{TASK_SLUG}}/research/{{TASK_SLUG}}-research.md`
+- **Template**: `{{RRCE_HOME}}/templates/research_output.md`
+- **Metadata**: `{{RRCE_DATA}}/tasks/{{TASK_SLUG}}/meta.json` with status `complete`
+- **Outcome**: Planning agent can proceed without re-asking the same questions
+
+## Integration Notes
+
+- **Init Agent**: Research works best after Init has established project context
+- **Planning Agent**: Receives research brief and creates execution plan
+- **Knowledge Base**: Research may identify new knowledge to document after implementation
+
+## Error Handling
+
+- **No project context**: Recommend `/init` but allow proceeding with limited context
+- **No search results**: Document that no prior work was found, proceed with clarification
+- **User unresponsive**: After 2 unanswered clarifications, document assumptions and offer to proceed or wait

package/agent-core/prompts/sync.md

@@ -1,8 +1,8 @@
 ---
 name: RRCE Sync
-description: Reconcile project state with the RRCE knowledge base.
-argument-hint: [SCOPE=<path|module>]
-tools: ['search/codebase']
+description: Reconcile project state with the RRCE knowledge base and update semantic index.
+argument-hint: "[SCOPE=<path|module>]"
+tools: ['search_knowledge', 'get_project_context', 'index_knowledge', 'list_projects']
 required-args: []
 optional-args:
   - name: SCOPE
@@ -14,13 +14,9 @@ auto-identity:
 
 You are the Knowledge Sync Lead. Act like a senior architect charged with keeping the RRCE knowledge cache authoritative and current.
 
-**⚠️ FIRST STEP (MANDATORY) - Path Resolution**
-Check if the system has pre-resolved paths for you. Look for a "System Resolved Paths" section at the start of this prompt context. If present, use those values directly:
-- `RRCE_DATA` = Pre-resolved data path (where knowledge, tasks, refs are stored)
-- `RRCE_HOME` = Pre-resolved global home
-- `WORKSPACE_ROOT` = Pre-resolved source code location
-
-**Only if no pre-resolved paths are present**, fall back to manual resolution by reading config.
+## Path Resolution
+Use the pre-resolved paths from the "System Resolved Paths" table in the context preamble.
+For details, see: `{{RRCE_HOME}}/docs/path-resolution.md`
 
 Pipeline Position
 - **Maintenance Agent**: Sync runs periodically or after significant codebase changes to keep knowledge current.
@@ -45,15 +41,6 @@ Non-Negotiables
 4. Keep all knowledge files lean (<500 lines each) and focused on durable insights, linking to code paths or task artifacts instead of duplicating detail.
 5. Record gaps or follow-up items in a checklist inside the file you touched so future runs can close them.
 
-Path Variables Reference
-- `{{RRCE_DATA}}` = Primary data path (knowledge, tasks, refs storage)
-- `{{RRCE_HOME}}` = Global RRCE home directory
-- `{{WORKSPACE_ROOT}}` = Source code directory
-- `{{WORKSPACE_NAME}}` = Project name
-
-Cross-Project References
-- Reference another project's context: `{{RRCE_HOME}}/workspaces/<other-project>/knowledge/`
-
 Workflow
 1. Review `{{RRCE_DATA}}/tasks/` and recent git history to identify areas that may have drifted from documented knowledge, prioritizing any scope passed via `SCOPE`.
 2. Inventory existing knowledge files. Note candidates for removal or consolidation when their scope is redundant or obsolete.
@@ -63,9 +50,11 @@ Workflow
    - Remove outdated sections or entire files once you verify the information no longer applies.
 4. Ensure cross-references (links to tasks, commits, or other knowledge files) point to current resources.
 5. Summarize any unresolved questions or future sync needs at the bottom of the modified file(s) under a `Checklist` heading.
-6. **Semantic Indexing**: If any knowledge files were modified or created, and the `index_knowledge` tool is available, run it to keep the search index current:
-   - Tool: `index_knowledge`
-   - Args: `{ project: "{{WORKSPACE_NAME}}" }`
+6. **Semantic Indexing (MANDATORY)**: After updating any knowledge files, run the indexer to keep search current:
+   ```
+   Tool: index_knowledge
+   Args: { "project": "{{WORKSPACE_NAME}}" }
+   ```
 
 Deliverable
 - Updated `{{RRCE_DATA}}/knowledge/*` files that accurately reflect the present project state, each carrying the latest `Updated:` marker and lean checklist.

package/agent-core/templates/doctor_output.md

@@ -1,11 +1,22 @@
 <!--
-  RRCE Template Variables:
-  - {{RRCE_DATA}}: Primary storage path (resolves based on storage mode in .rrce-workflow/config.yaml)
-      - global: {{RRCE_HOME}}/workspaces/<workspace-name>/
-      - workspace: <workspace>/.rrce-workflow/
-  - {{RRCE_HOME}}: Global home (default: ~/.rrce-workflow, customizable via storage.globalPath in config)
-  - {{WORKSPACE_ROOT}}: Workspace root directory
-  - {{WORKSPACE_NAME}}: Workspace name from config or directory name
+  TEMPLATE: Project Diagnosis
+  
+  HOW TO USE:
+  1. Copy to: {{RRCE_DATA}}/tasks/doctor-{{YYYYMMDD}}/diagnosis.md
+  2. Replace {{variable}} placeholders with actual values
+  3. Add findings under appropriate priority sections
+  4. Remove empty sections and placeholder rows
+  
+  AUTO-FILLED VARIABLES (from System Resolved Paths):
+  - {{RRCE_DATA}}: Storage path for knowledge/tasks
+  - {{WORKSPACE_ROOT}}: Source code directory
+  
+  AGENT-FILLED VARIABLES:
+  - {{project_name}}: Name of the project
+  - {{date}}: ISO date (YYYY-MM-DD)
+  - {{author}}: Git user or agent name (analyst)
+  - {{workspace_root}}: Source code directory
+  - {{focus_area}}: Focus area parameter or "General"
 -->
 # Project Diagnosis – {{project_name}}
 

package/agent-core/templates/documentation_output.md

@@ -1,11 +1,27 @@
 <!--
-  RRCE Template Variables:
-  - {{RRCE_DATA}}: Primary storage path (resolves based on storage mode in .rrce-workflow/config.yaml)
-      - global: {{RRCE_HOME}}/workspaces/<workspace-name>/
-      - workspace: <workspace>/.rrce-workflow/
-  - {{RRCE_HOME}}: Global home (default: ~/.rrce-workflow, customizable via storage.globalPath in config)
-  - {{WORKSPACE_ROOT}}: Workspace root directory
-  - {{WORKSPACE_NAME}}: Workspace name from config or directory name
+  TEMPLATE: Handover Note / Documentation
+  
+  HOW TO USE:
+  1. Copy to destination based on DOC_TYPE:
+     - With TASK_SLUG: {{RRCE_DATA}}/tasks/{{TASK_SLUG}}/docs/{{TASK_SLUG}}-{{DOC_TYPE}}.md
+     - With TARGET_PATH: {{RRCE_DATA}}/{{TARGET_PATH}}
+     - Default: {{RRCE_DATA}}/knowledge/{{DOC_TYPE}}.md
+  2. Replace {{variable}} placeholders with actual values
+  3. Remove empty sections
+  
+  AUTO-FILLED VARIABLES (from System Resolved Paths):
+  - {{RRCE_DATA}}: Storage path for knowledge/tasks
+  - {{WORKSPACE_NAME}}: Project name
+  
+  AGENT-FILLED VARIABLES:
+  - {{task_id}}: UUID for the task (if applicable)
+  - {{task_slug}}: kebab-case task identifier (if applicable)
+  - {{task_title}}: Human-readable task title
+  - {{author}}: Git user or agent name
+  - {{date}}: ISO date (YYYY-MM-DD)
+  - {{execution_artifact}}: Path to execution log
+  - {{release_ref}}: Release tag or commit SHA
+  - {{workspace_name}}: Project name
 -->
 # Handover Note – {{task_title}}
 

package/agent-core/templates/executor_output.md

@@ -1,11 +1,25 @@
 <!--
-  RRCE Template Variables:
-  - {{RRCE_DATA}}: Primary storage path (resolves based on storage mode in .rrce-workflow/config.yaml)
-      - global: {{RRCE_HOME}}/workspaces/<workspace-name>/
-      - workspace: <workspace>/.rrce-workflow/
-  - {{RRCE_HOME}}: Global home (default: ~/.rrce-workflow, customizable via storage.globalPath in config)
-  - {{WORKSPACE_ROOT}}: Workspace root directory
-  - {{WORKSPACE_NAME}}: Workspace name from config or directory name
+  TEMPLATE: Execution Log
+  
+  HOW TO USE:
+  1. Copy to: {{RRCE_DATA}}/tasks/{{TASK_SLUG}}/execution/{{TASK_SLUG}}-execution.md
+  2. Replace {{variable}} placeholders with actual values
+  3. Update Implementation Steps table as work progresses
+  4. Remove empty sections
+  
+  AUTO-FILLED VARIABLES (from System Resolved Paths):
+  - {{RRCE_DATA}}: Storage path for knowledge/tasks
+  - {{WORKSPACE_NAME}}: Project name
+  
+  AGENT-FILLED VARIABLES:
+  - {{task_id}}: UUID for the task
+  - {{task_slug}}: kebab-case task identifier
+  - {{task_title}}: Human-readable task title
+  - {{author}}: Git user or agent name (executor)
+  - {{date}}: ISO date (YYYY-MM-DD)
+  - {{plan_artifact}}: Path to execution plan
+  - {{git_ref}}: Branch name or commit SHA
+  - {{workspace_name}}: Project name
 -->
 # Execution Log – {{task_title}}
 

package/agent-core/templates/init_output.md

@@ -1,11 +1,31 @@
 <!--
-  RRCE Template Variables:
-  - {{RRCE_DATA}}: Primary storage path (resolves based on storage mode in .rrce-workflow/config.yaml)
-      - global: {{RRCE_HOME}}/workspaces/<workspace-name>/
-      - workspace: <workspace>/.rrce-workflow/
-  - {{RRCE_HOME}}: Global home (default: ~/.rrce-workflow, customizable via storage.globalPath in config)
-  - {{WORKSPACE_ROOT}}: Workspace root directory
-  - {{WORKSPACE_NAME}}: Workspace name from config or directory name
+  TEMPLATE: Project Context (Init Output)
+  
+  HOW TO USE:
+  1. Copy to: {{RRCE_DATA}}/knowledge/project-context.md
+  2. Replace {{variable}} placeholders with actual values
+  3. Delete sections marked (OPTIONAL) if empty after population
+  4. Fill in remaining sections with discovered information
+  
+  AUTO-FILLED VARIABLES (from System Resolved Paths):
+  - {{WORKSPACE_ROOT}}: Source code directory
+  - {{WORKSPACE_NAME}}: Project name  
+  - {{RRCE_DATA}}: Storage path for knowledge/tasks
+  - {{RRCE_HOME}}: Global RRCE home directory
+  
+  AGENT-FILLED VARIABLES:
+  - {{project_name}}: Name of the project
+  - {{date}}: ISO date (YYYY-MM-DD)
+  - {{author}}: Git user or agent name
+  - {{workspace_root}}: Same as WORKSPACE_ROOT
+  
+  DYNAMIC VARIABLES:
+  - {{test_unit_command}}: Command to run unit tests
+  - {{test_integration_command}}: Command to run integration tests
+  - {{test_e2e_command}}: Command to run e2e tests
+  - {{coverage_command}}: Command to generate coverage report
+  - {{last_indexed_date}}: Last semantic index update
+  - {{index_path}}: Path to embeddings.json
 -->
 # Project Context – {{project_name}}
 

package/agent-core/templates/planning_output.md

@@ -1,11 +1,23 @@
 <!--
-  RRCE Template Variables:
-  - {{RRCE_DATA}}: Primary storage path (resolves based on storage mode in .rrce-workflow/config.yaml)
-      - global: {{RRCE_HOME}}/workspaces/<workspace-name>/
-      - workspace: <workspace>/.rrce-workflow/
-  - {{RRCE_HOME}}: Global home (default: ~/.rrce-workflow, customizable via storage.globalPath in config)
-  - {{WORKSPACE_ROOT}}: Workspace root directory
-  - {{WORKSPACE_NAME}}: Workspace name from config or directory name
+  TEMPLATE: Execution Plan
+  
+  HOW TO USE:
+  1. Copy to: {{RRCE_DATA}}/tasks/{{TASK_SLUG}}/planning/{{TASK_SLUG}}-plan.md
+  2. Replace {{variable}} placeholders with actual values
+  3. Remove unused table rows and empty sections
+  
+  AUTO-FILLED VARIABLES (from System Resolved Paths):
+  - {{RRCE_DATA}}: Storage path for knowledge/tasks
+  - {{WORKSPACE_NAME}}: Project name
+  
+  AGENT-FILLED VARIABLES:
+  - {{task_id}}: UUID for the task
+  - {{task_slug}}: kebab-case task identifier
+  - {{task_title}}: Human-readable task title
+  - {{author}}: Git user or agent name (planner)
+  - {{date}}: ISO date (YYYY-MM-DD)
+  - {{research_artifact}}: Path to research brief
+  - {{workspace_name}}: Project name
 -->
 # Execution Plan – {{task_title}}
 

package/agent-core/templates/research_output.md

@@ -1,11 +1,23 @@
 <!--
-  RRCE Template Variables:
-  - {{RRCE_DATA}}: Primary storage path (resolves based on storage mode in .rrce-workflow/config.yaml)
-      - global: {{RRCE_HOME}}/workspaces/<workspace-name>/
-      - workspace: <workspace>/.rrce-workflow/
-  - {{RRCE_HOME}}: Global home (default: ~/.rrce-workflow, customizable via storage.globalPath in config)
-  - {{WORKSPACE_ROOT}}: Workspace root directory
-  - {{WORKSPACE_NAME}}: Workspace name from config or directory name
+  TEMPLATE: Research Brief
+  
+  HOW TO USE:
+  1. Copy to: {{RRCE_DATA}}/tasks/{{TASK_SLUG}}/research/{{TASK_SLUG}}-research.md
+  2. Replace {{variable}} placeholders with actual values
+  3. Delete sections that are empty after population
+  
+  AUTO-FILLED VARIABLES (from System Resolved Paths):
+  - {{RRCE_DATA}}: Storage path for knowledge/tasks
+  - {{WORKSPACE_NAME}}: Project name
+  
+  AGENT-FILLED VARIABLES:
+  - {{task_id}}: UUID for the task
+  - {{task_slug}}: kebab-case task identifier
+  - {{task_title}}: Human-readable task title
+  - {{author}}: Git user or agent name
+  - {{date}}: ISO date (YYYY-MM-DD)
+  - {{source}}: URL or reference to original request
+  - {{workspace_name}}: Project name
 -->
 # Research Brief – {{task_title}}