investigation-workflow

Relationship to Dev Orchestrator

Normal execution path: This workflow is invoked as a sub-recipe by the dev-orchestrator skill via smart-orchestrator. You do NOT normally need to activate this skill directly.

User request → dev-orchestrator → smart-orchestrator recipe
    → investigation-workflow recipe (this skill's recipe)

Direct invocation is supported as a compatibility path when the dev-orchestrator is unavailable or when explicitly requested. In that case, use the recipe runner (see Execution Instructions below).

Workflow Graph

flowchart TD
    INIT[Initialize Tracking] --> P1

    subgraph P1["Phase 1: Scope Definition"]
        SCOPE[scope-definition<br/>prompt-writer agent] --> AMB{Has ambiguities?}
        AMB -->|yes| CLARIFY[clarify-ambiguities<br/>ambiguity agent]
        AMB -->|no| P1_OUT[Scope defined]
        CLARIFY --> P1_OUT
    end

    subgraph P2["Phase 2: Exploration Strategy"]
        STRAT[exploration-strategy<br/>architect agent] --> PAST[check-past-investigations<br/>patterns agent]
        PAST --> HIST{Historical context needed?}
        HIST -->|yes| ARCH[historical-research<br/>knowledge-archaeologist]
        HIST -->|no| P2_OUT[Strategy ready]
        ARCH --> P2_OUT
    end

    subgraph P3["Phase 3: Parallel Deep Dives"]
        DD1[deep-dive-primary<br/>architect agent]
        DD2[deep-dive-secondary<br/>patterns agent]
        DD3[deep-dive-tertiary<br/>architect agent]
        DD4{Specialist needed?}
        DD4 -->|yes| DDS[deep-dive-specialist<br/>security agent]
        DD1 & DD2 & DD3 --> CONSOL[consolidate-findings<br/>patterns agent]
        DDS --> CONSOL
    end

    subgraph P4["Phase 4: Verification"]
        HYP[formulate-hypotheses<br/>architect agent] --> EXEC[execute-verification<br/>architect agent]
        EXEC --> VAL[validate-verification<br/>reviewer agent]
    end

    subgraph P5["Phase 5: Synthesis"]
        PAT[identify-patterns<br/>patterns agent] --> SYN[synthesis<br/>architect agent]
        SYN --> VSYN[validate-synthesis<br/>reviewer agent]
    end

    subgraph P6["Phase 6: Knowledge Capture"]
        DISC[update-discoveries] --> PATN{New patterns?}
        PATN -->|yes| UPAT[update-patterns]
        PATN -->|no| RPT[create-investigation-report]
        UPAT --> RPT
    end

    P1 --> P2 --> P3 --> P4 --> P5 --> P6

    RPT --> TRANS[transition-guidance<br/>patterns agent]
    TRANS --> EFF[efficiency-report]
    EFF --> FINAL[final-output]

    TRANS --> TDEV{Transition to dev?}
    TDEV -->|yes| DW[Launch default-workflow<br/>recipe via recipe runner]
    TDEV -->|no| DONE[Investigation Complete]

Purpose

This skill provides a systematic 6-phase workflow for investigating and understanding existing systems, codebases, and architectures. Unlike development workflows optimized for implementation, this workflow is optimized for exploration, understanding, and knowledge capture.

It is normally executed as a sub-recipe by the dev-orchestrator via smart-orchestrator, but can also be invoked directly via the recipe runner.

Canonical Sources

• Executable source (recipe): amplifier-bundle/recipes/investigation-workflow.yaml
• Reference documentation: .claude/workflow/INVESTIGATION_WORKFLOW.md

The recipe YAML is the authoritative execution definition. The .md file serves as human-readable reference documentation for the workflow phases.

Execution Instructions

Normal path (via dev-orchestrator)

If you reached this skill via dev-orchestrator / smart-orchestrator, the recipe runner is already managing execution. Do not re-invoke the recipe runner. The orchestrator handles the full lifecycle including goal-seeking reflection loops.

Direct invocation (standalone)

If this skill is activated directly (not via dev-orchestrator), you MUST use the recipe runner — do NOT read the .md file and follow phases manually:

from amplihack.recipes import run_recipe_by_name

result = run_recipe_by_name(
    "investigation-workflow",
    user_context={
        "task_description": "TASK_DESCRIPTION_HERE",
        "repo_path": ".",
    },
    progress=True,
)

Or via shell:

cd /path/to/repo && env -u CLAUDECODE \
  AMPLIHACK_HOME=/path/to/amplihack PYTHONPATH=${AMPLIHACK_HOME:-~/.amplihack}/src python3 -c "
from amplihack.recipes import run_recipe_by_name
result = run_recipe_by_name('investigation-workflow', user_context={
    'task_description': '''TASK_DESCRIPTION_HERE''',
    'repo_path': '.',
}, progress=True)
print(f'Recipe result: {result}')
"

Do NOT read INVESTIGATION_WORKFLOW.md and follow phases manually. The recipe runner enforces phase ordering, agent deployment, and quality gates that manual execution cannot replicate.

Preferred: Use dev-orchestrator instead

For most tasks, invoke Skill(skill="dev-orchestrator") or use /dev <task> rather than activating this skill directly. The dev-orchestrator adds goal-seeking reflection, workstream decomposition, and adaptive error recovery on top of this workflow.

When to Use This Skill

Investigation Tasks (use this workflow):

• "Investigate how the authentication system works"
• "Explain the neo4j memory integration"
• "Understand why CI is failing consistently"
• "Analyze the reflection system architecture"
• "Research what hooks are triggered during session start"

Development Tasks (use default-workflow recipe instead):

• "Implement OAuth support"
• "Build a new API endpoint"
• "Add feature X"
• "Fix bug Y"

Core Philosophy

Exploration First: Define scope and strategy before diving into code Parallel Deep Dives: Deploy multiple agents simultaneously for efficient information gathering Verification Required: Test understanding through practical application Knowledge Capture: Document findings to prevent repeat investigations

The 6-Phase Investigation Workflow

Phase 1: Scope Definition

Purpose: Define investigation boundaries and success criteria before any exploration.

Tasks:

• FIRST: Identify explicit user requirements - What specific questions must be answered?
• Use prompt-writer agent to clarify investigation scope
• Use ambiguity agent if questions are unclear
• Define what counts as "understanding achieved"
• List specific questions that must be answered
• Set boundaries: What's in scope vs. out of scope
• Estimate investigation depth needed (surface-level vs. deep dive)

Phase 2: Exploration Strategy

Purpose: Plan which agents to deploy and what to investigate, preventing inefficient random exploration.

Tasks:

• Use architect agent to design exploration strategy
• Use patterns agent to check for similar past investigations
• Identify key areas to explore (code paths, configurations, documentation)
• Select specialized agents for parallel deployment in Phase 3

Phase 3: Parallel Deep Dives

Purpose: Deploy multiple exploration agents simultaneously to gather information efficiently.

CRITICAL: This phase uses PARALLEL EXECUTION by default.

Phase 4: Verification & Testing

Purpose: Test and validate understanding through practical application.

Phase 5: Synthesis

Purpose: Compile findings into coherent explanation that answers original questions.

Phase 6: Knowledge Capture

Purpose: Create durable documentation so this investigation never needs to be repeated.

• Store discoveries in memory using store_discovery() from amplihack.memory.discoveries
• Update .claude/context/PATTERNS.md if reusable patterns found

Transitioning to Development Workflow

After investigation completes, if the task requires implementation, the dev-orchestrator handles the transition automatically via its goal-seeking reflection loop. If running standalone, transition by launching the default-workflow recipe:

run_recipe_by_name("default-workflow", user_context={
    "task_description": "Implement findings from investigation...",
    "repo_path": ".",
}, progress=True)

Integration with Dev Orchestrator

The dev-orchestrator automatically detects investigation tasks using keywords and routes them to this workflow's recipe:

User: "/dev investigate how authentication works"

dev-orchestrator: Classified as Investigation → launching investigation-workflow recipe
→ Recipe runner executes 6-phase investigation workflow
→ Results feed into goal-seeking reflection loop

For hybrid tasks (investigate + implement), the dev-orchestrator decomposes into parallel workstreams: one running investigation-workflow, another running default-workflow.

Key Principles

• Scope first, explore second - Define boundaries before diving in
• Parallel exploration is key - Deploy multiple agents simultaneously in Phase 3
• Verify understanding - Test your hypotheses in Phase 4
• Capture knowledge - Always store discoveries in memory in Phase 6
• This workflow optimizes for understanding, not implementation

Related Resources

• Recipe (executable): amplifier-bundle/recipes/investigation-workflow.yaml
• Reference docs: .claude/workflow/INVESTIGATION_WORKFLOW.md
• Dev Orchestrator: .claude/skills/dev-orchestrator/
• Default Workflow: .claude/skills/default-workflow/
• Agent Catalog: .claude/agents/amplihack/ directory
• Pattern Library: .claude/context/PATTERNS.md