Spec Driven Analytics

Spec-Driven Analytics

Orchestrate session data mining operations through the session-miner subagent to deliver formatted, actionable analytics from Claude Code session history, with 4-layer structural anti-skip enforcement.

Context files are THE LAW: tech-stack.md, source-tree.md, dependencies.md, coding-standards.md, architecture-constraints.md, anti-patterns.md

If ambiguous or conflicts detected: HALT and use AskUserQuestion

Execution Model

This skill expands inline. After invocation, execute Phase 00 Initialization immediately. Do not wait passively, ask permission, or offer execution options.

Self-Check (if ANY box is true = VIOLATION):

• Stopping to ask about token budget
• Stopping to offer execution options
• Waiting passively for user to say "go"
• Asking "should I execute this?"
• Skipping a phase because it "seems simple"
• Combining multiple phases into one
• Summarizing instead of loading a reference file
• Skipping verification because "I already wrote the file"

IF any box checked: EXECUTION MODEL VIOLATION. Go directly to Phase 00 Initialization now.

Anti-Skip Enforcement Contract

This skill enforces 4 independent anti-skip layers. ALL FOUR must fail for a step to be skipped:

1. Per-phase reference loading - Each phase loads its reference files fresh via Read(). NOT consolidated. Prevents "already covered" rationalization.
2. Binary CLI gates - devforgeai-validate phase-check/phase-complete at phase boundaries. Cannot be forged by LLM.
3. Checkpoint-based state tracking - Phase completion verified by checking checkpoint JSON data keys and current_phase field.
4. Artifact verification - Checkpoint JSON existence checked via Glob(), cache files verified on disk, output verified via Grep().

Execute-Verify-Record Pattern: Every mandatory step in every phase file has three parts:

• EXECUTE: The exact action to perform (Read, Write, Task, Glob, Grep, AskUserQuestion)
• VERIFY: How to confirm the action happened (file exists, content contains expected text, data key populated)
• RECORD: Update checkpoint JSON with captured data; call devforgeai-validate phase-record

Token Optimization Bias is PROHIBITED. Do not skip, compress, or shortcut any step. Every phase step exists because a previous failure proved it necessary.

Purpose

This skill serves as the analytics orchestration layer for DevForgeAI. It coordinates the session-miner subagent to extract patterns from Claude Code history.jsonl files and presents formatted, actionable analytics.

Core Philosophy

"Mine Once, Reference Forever" - Extract session patterns into cached analytics that survive context window clears and are queryable across sessions.

"Evidence-Based Optimization" - All workflow recommendations are backed by empirical session data, not speculation.

"Data Compounds Over Time" - Analytics cache grows with each query, related patterns are cross-referenced, and staleness tracking ensures freshness.

When to Use This Skill

Trigger Scenarios

• Analyzing workflow patterns and execution frequencies
• Mining error patterns and failure points from session history
• Surfacing architectural and implementation decisions
• Deep analysis of a specific story's development history
• Identifying high-frequency command sequences for optimization
• User runs /analytics command
• User mentions session analytics, workflow patterns, or error mining

When NOT to Use

• Capturing feedback (use spec-driven-feedback)
• Searching chat history (use /chat-search)
• Root cause analysis of specific failures (use spec-driven-rca)

Parameter Extraction

Extract from command arguments:

Default: If no arguments, $QUERY_TYPE = "dashboard"

State Persistence

• Checkpoint: devforgeai/workflows/${ANALYTICS_ID}-phase-state.json
• References: references/ (self-contained within this skill)
• Templates: assets/templates/ (self-contained within this skill)
• Cache Output: devforgeai/cache/analytics/

Phase 00: Initialization [INLINE - Bootstraps State]

This phase runs inline because it creates the state that all other phases depend on. It CANNOT be skipped.

Step 0.1: Parse Arguments

Extract arguments from command invocation. Determine mode:

IF args contain "--help" or "-h":
  Read("src/claude/skills/spec-driven-analytics/references/analytics-help.md")
  Display help content
  EXIT skill

IF args contain "--resume ANALYTICS-NNN":
  mode = "resume"
  RESUME_ID = extract_id(args)
  GOTO Step 0.2 (Resume Detection)

# Parse query type
QUERY_TYPE = first_positional_arg OR "dashboard"
VALID_TYPES = ["dashboard", "workflows", "errors", "decisions", "story", "command-patterns"]

IF QUERY_TYPE not in VALID_TYPES:
  Display error: "Invalid query type: '{QUERY_TYPE}'"
  Display: "Valid types: dashboard, workflows, errors, decisions, story, command-patterns"
  HALT

# Parse optional parameters
STORY_ID = extract_story_id(args)  # Required if QUERY_TYPE == "story"
FORCE_REFRESH = "--force" in args
DAYS_LIMIT = extract_days(args)    # Default: null (all data)
QUERY_PARAM = extract_quoted_string(args)  # For decisions query

IF QUERY_TYPE == "story" AND STORY_ID is null:
  Display error: "Story query requires a STORY-ID parameter"
  Display: "Usage: /analytics story STORY-XXX"
  HALT

Step 0.2: Resume Detection

IF mode == "resume":
  checkpoint_file = Glob(pattern=f"devforgeai/workflows/{RESUME_ID}-phase-state.json")
  IF checkpoint_file found:
    Read the checkpoint file
    Restore state from checkpoint
    Set CURRENT_PHASE from checkpoint.progress.current_phase
    GOTO Phase Orchestration Loop at CURRENT_PHASE
  ELSE:
    HALT -- "Analytics session {RESUME_ID} not found"

# Check for existing in-progress checkpoint (new mode)
existing_checkpoints = Glob(pattern="devforgeai/workflows/ANALYTICS-*-phase-state.json")
IF existing_checkpoints with status "in_progress":
  AskUserQuestion:
    Question: "Found existing analytics session in progress. Resume or start fresh?"
    Header: "Resume"
    Options:
      - label: "Resume session"
        description: "Continue from last checkpoint"
      - label: "Start fresh"
        description: "Begin new analytics session"
  IF "Resume": Restore state, GOTO Phase Orchestration Loop at CURRENT_PHASE

Step 0.3: Generate Analytics ID

existing_files = Glob("devforgeai/workflows/ANALYTICS-*-phase-state.json")

# Extract date-based IDs (gap-aware)
today = current_date_YYYY_MM_DD
today_checkpoints = [f for f in existing_files if today in f]
next_seq = len(today_checkpoints) + 1

ANALYTICS_ID = f"ANALYTICS-{today}-{next_seq:03d}"

Step 0.4: CLI Initialization

source .venv/bin/activate && devforgeai-validate phase-init ${ANALYTICS_ID} --workflow=analytics --project-root=. 2>&1

Step 0.5: Create Initial Checkpoint

{
  "checkpoint_version": "1.0",
  "analytics_id": "${ANALYTICS_ID}",
  "workflow": "analytics",
  "created_at": "ISO 8601",
  "updated_at": "ISO 8601",
  "status": "in_progress",
  "input": {
    "query_type": "${QUERY_TYPE}",
    "story_id": "${STORY_ID}",
    "force_refresh": false,
    "days_limit": null,
    "query_param": null
  },
  "progress": {
    "current_phase": 0,
    "phases_completed": [],
    "total_steps_completed": 0
  },
  "phases": {
    "01": { "status": "pending", "steps_completed": [] },
    "02": { "status": "pending", "steps_completed": [] },
    "03": { "status": "pending", "steps_completed": [] },
    "04": { "status": "pending", "steps_completed": [] },
    "05": { "status": "pending", "steps_completed": [] },
    "06": { "status": "pending", "steps_completed": [] }
  },
  "cache": {
    "cache_key": null,
    "cache_hit": false,
    "cache_path": null
  },
  "output": {
    "raw_entries": null,
    "aggregated_results": null,
    "formatted_output": null,
    "error": null
  }
}

Write to devforgeai/workflows/${ANALYTICS_ID}-phase-state.json

VERIFY: Glob(pattern="devforgeai/workflows/${ANALYTICS_ID}-phase-state.json") IF not found: HALT -- "Initial checkpoint was NOT created."

Step 0.6: Display Session Banner

Display:
"------------------------------------------------------------
  DevForgeAI Analytics Session
------------------------------------------------------------

Analytics ID: ${ANALYTICS_ID}
Query Type:   ${QUERY_TYPE}
Story ID:     ${STORY_ID || 'N/A'}
Force Refresh: ${FORCE_REFRESH}
Days Limit:   ${DAYS_LIMIT || 'All data'}

Phases: 7 (Initialization > Cache Management > Query Orchestration > Result Aggregation > Output Formatting > Display & Delivery > Completion)
------------------------------------------------------------"

Step 0.7: Checkpoint Path Verification

Glob(pattern="devforgeai/workflows/${ANALYTICS_ID}-phase-state.json")
IF not found: HALT -- "Checkpoint file missing after Step 0.6"

Set CURRENT_PHASE = 1.

Phase Orchestration Loop

phase_slugs = {
    1: "cache-management",
    2: "query-orchestration",
    3: "result-aggregation",
    4: "output-formatting",
    5: "display-delivery",
    6: "completion-summary"
}

FOR phase_num in range(CURRENT_PHASE, 7):  # Phases 01-06
    phase_id = f"{phase_num:02d}"
    slug = phase_slugs[phase_num]

    1. ENTRY GATE: devforgeai-validate phase-check ${ANALYTICS_ID} --workflow=analytics --from={prev_phase} --to={phase_id} --project-root=.
       IF exit != 0 AND exit != 127: HALT

    2. LOAD: Read(file_path=f"src/claude/skills/spec-driven-analytics/phases/phase-{phase_id}-{slug}.md")
       Load the phase file FRESH. Do NOT skip this step. Do NOT rely on memory of previous reads.

    3. REFERENCE: Read the phase's reference files as specified in the phase Contract section.
       References are in references/ (self-contained within this skill).
       Load ALL listed references. Do not skip any.

    4. EXECUTE: Follow EVERY step in the phase file using EXECUTE-VERIFY-RECORD triplets.

    5. EXIT GATE: devforgeai-validate phase-complete ${ANALYTICS_ID} --workflow=analytics --phase={phase_id} --checkpoint-passed --project-root=.
       IF exit != 0 AND exit != 127: HALT

    6. CHECKPOINT: Update checkpoint JSON with phase completion.
       Write updated checkpoint to disk.
       Verify write via Glob().

    # CACHE HIT SHORT-CIRCUIT: If Phase 01 sets CACHE_HIT=true, skip Phases 02-04 and jump to Phase 05.
    IF phase_num == 1 AND CACHE_HIT == true:
       Set CURRENT_PHASE = 5
       CONTINUE to Phase 05 (Display & Delivery)

Phase Table

Required Subagents Per Phase

All other phases: No subagents required. Direct tool calls (Read, Write, Glob, Grep, AskUserQuestion).

Deviation Protocol: Any skip requires explicit user consent via AskUserQuestion.

Workflow Completion Validation

completed_count = len(checkpoint.progress.phases_completed)
IF completed_count < 6:
    HALT "WORKFLOW INCOMPLETE - {completed_count}/6 phases completed"
IF completed_count == 6:
    Display "All 6 phases completed - Workflow validation passed"
    Update checkpoint status to "completed"

Success Criteria

Complete analytics query processed with:

• Valid analytics ID (ANALYTICS-YYYY-MM-DD-NNN format)
• Cache checked (hit returns immediately, miss proceeds to mining)
• session-miner subagent invoked via Task() pattern (on cache miss)
• Results aggregated, filtered, and ranked
• Output formatted as markdown with tables
• Cache mechanism operational (1-hour TTL)
• All 6 query types supported (dashboard, workflows, errors, decisions, story, command-patterns)
• Checkpoint updated to "completed"

Reference Files Inventory

Load these on-demand during workflow execution:

Phase Files (6 files in phases/)

Reference Files (6 files in references/)

• cache-management.md - TTL logic, cache key generation, invalidation rules, force refresh behavior
• query-configuration.md - 6 query type configurations, prompt templates, parameter mapping
• session-miner-delegation.md - Subagent invocation contract, Task() template, response schema
• aggregation-pipeline.md - group_by, filter_by, calculate_metrics, rank_by_relevance
• output-templates.md - 6 query-specific markdown templates, table schemas
• analytics-help.md - Help text, usage examples, error handling

Assets (1 template)

• assets/templates/checkpoint-template.json - Checkpoint schema template

Total: 6 phase files + 6 reference files + 1 template = 13 files + SKILL.md = 14 files

Error Handling

Subagent Errors

IF session-miner returns error:
    Log error details to checkpoint.output.error
    Return error template with troubleshooting steps
    Do NOT cache error responses

Cache Errors

IF cache read fails:
    Proceed without cache (CACHE_HIT=false)
    Log warning in checkpoint

Empty Results

IF no matching sessions found:
    Return "no results" template
    Suggest query modifications via AskUserQuestion

Best Practices

1. Use specific query types - Dashboard for overview, targeted queries for deep dives
2. Use --force sparingly - Cache exists to save time; force-refresh when data may have changed
3. Use --days for recent analysis - Narrow time windows produce faster, more focused results
4. Review recommendations - Analytics recommendations are data-driven but require human judgment
5. Link to stories/ADRs when relevant - Bidirectional traceability strengthens the knowledge base

See phase-specific reference files for detailed procedures.