tools-only Master Workflow Overview
Complete visualization of the 6-stage agentic workflow with all repository assets mapped.
High-Level Flow
flowchart TB
subgraph INPUT["1. INPUT RECEPTION ⚠️"]
I1[User Request]
I2["/sessions command"]
I3["session-start-rtica.cjs hook"]
end
subgraph CONTEXT["2. CONTEXT GATHERING ✅"]
C1["context-gathering agent"]
C2["context-refinement agent"]
C3["rt-ica skill"]
C4["Reference Skills<br/>(4 skills)"]
end
subgraph PLANNING["3. PLANNING ✅"]
P1["rt-ica skill"]
P2["delegate skill"]
P3["/how-to-delegate"]
P4["/think command"]
P5["scientific-thinking skill"]
end
subgraph EXECUTION["4. EXECUTION ✅"]
E1["Specialist Agents<br/>(6 agents)"]
E2["subagent-contract skill"]
E3["agent-creator skill"]
end
subgraph VERIFICATION["5. VERIFICATION ✅"]
V1["verify skill"]
V2["audit skill"]
V3["/am-i-complete"]
V4["code-review agent"]
V5["plugin-assessor agent"]
V6["doc-drift-auditor agent"]
end
subgraph OUTPUT["6. OUTPUT DELIVERY ⚠️"]
O1["git-commit-helper skill"]
O2["DONE/BLOCKED signaling"]
O3["plugin-docs-writer agent"]
O4["logging agent"]
end
INPUT --> CONTEXT
CONTEXT --> PLANNING
PLANNING --> EXECUTION
EXECUTION --> VERIFICATION
VERIFICATION --> OUTPUT
VERIFICATION -.->|"Issues found"| EXECUTION
Detailed Stage Breakdown
Stage 1: Input Reception ⚠️ PARTIAL
flowchart LR
subgraph TRIGGER["Triggers"]
T1["User message"]
T2["Session start"]
T3["Hook execution"]
end
subgraph ASSETS["Available Assets"]
A1["/sessions<br/>Session management"]
A2["session-start-rtica.cjs<br/>Auto-trigger rt-ica"]
end
subgraph GAPS["Identified Gaps"]
G1["❌ Task classification"]
G2["❌ Request routing"]
G3["❌ Priority assignment"]
end
T1 --> A1
T2 --> A2
T3 --> A2
Coverage Details:
| Asset | Type | Function |
|---|---|---|
| /sessions | Command | Manage session state, trigger modes |
| session-start-rtica.cjs | Hook | Auto-load rt-ica on session start |
Gap Analysis: No automated classification of incoming requests. Manual determination required for task type, complexity, and routing.
Stage 2: Context Gathering ✅ COVERED
flowchart TB
subgraph RESEARCH["Research Phase"]
R1["context-gathering agent<br/>Deep codebase research"]
R2["context-refinement agent<br/>Update context manifest"]
end
subgraph PREREQ["Prerequisites Check"]
P1["rt-ica skill<br/>Requirements assessment"]
end
subgraph REFERENCE["Reference Loading"]
REF1["claude-skills-overview-2026"]
REF2["claude-plugins-reference-2026"]
REF3["claude-commands-reference-2026"]
REF4["claude-hooks-reference-2026"]
end
subgraph GAP["Gaps"]
G1["❌ Context caching"]
end
R1 --> R2
P1 --> R1
REFERENCE --> R1
Coverage Details:
| Asset | Type | Function |
|---|---|---|
| context-gathering | Agent | Research without polluting orchestrator context |
| context-refinement | Agent | Update task context manifest with discoveries |
| rt-ica | Skill | Verify prerequisites before planning |
| claude-skills-overview-2026 | Skill | Reference for skills system |
| claude-plugins-reference-2026 | Skill | Reference for plugins system |
| claude-commands-reference-2026 | Skill | Reference for commands system |
| claude-hooks-reference-2026 | Skill | Reference for hooks system |
Gap Analysis: No persistent caching of gathered context across sessions.
Stage 3: Planning ✅ COVERED
flowchart TB
subgraph ASSESS["Assessment"]
A1["rt-ica skill<br/>Information completeness"]
end
subgraph DECOMPOSE["Decomposition"]
D1["/think command<br/>Step-back reasoning"]
D2["scientific-thinking skill<br/>Hypothesis formation"]
end
subgraph DELEGATE["Delegation Design"]
DEL1["delegate skill<br/>WHERE-WHAT-WHY template"]
DEL2["/how-to-delegate<br/>Full delegation guidance"]
end
subgraph GAP["Gaps"]
G1["❌ Complexity estimation"]
G2["❌ Time/effort prediction"]
end
A1 --> D1
D1 --> D2
D2 --> DEL1
DEL1 --> DEL2
Coverage Details:
| Asset | Type | Function |
|---|---|---|
| rt-ica | Skill | Block planning until prerequisites verified |
| delegate | Skill | Quick WHERE-WHAT-WHY template |
| /how-to-delegate | Command | Comprehensive delegation framework |
| /think | Command | Step-back broader perspective |
| scientific-thinking | Skill | Hypothesis-driven approach |
| /step-back | Command | Wider view of task implications |
| /scientific-thinking | Command | Activate scientific method |
Gap Analysis: No automated complexity scoring or effort estimation.
Stage 4: Execution ✅ COVERED
flowchart TB
subgraph SPECIALISTS["Specialist Agents"]
S1["skill-refactorer<br/>Split large skills"]
S2["claude-context-optimizer<br/>Improve AI docs"]
S3["subagent-refactorer<br/>Improve agent prompts"]
S4["subagent-generator<br/>Create new agents"]
S5["plugin-docs-writer<br/>Generate documentation"]
S6["context-gathering<br/>Research tasks"]
end
subgraph GOVERNANCE["Execution Governance"]
G1["subagent-contract skill<br/>Role boundaries"]
G2["DONE/BLOCKED signaling"]
end
subgraph CREATE["Agent Creation"]
C1["agent-creator skill<br/>Build new specialists"]
end
subgraph GAP["Gaps"]
GAP1["❌ Error recovery"]
GAP2["❌ Rollback mechanism"]
GAP3["❌ Retry logic"]
end
G1 --> SPECIALISTS
SPECIALISTS --> G2
C1 --> SPECIALISTS
Coverage Details:
| Asset | Type | Function |
|---|---|---|
| skill-refactorer | Agent | Refactor large skills into focused units |
| claude-context-optimizer | Agent | Optimize AI-facing documentation |
| subagent-refactorer | Agent | Improve agent prompt quality |
| subagent-generator | Agent | Create new agent definitions |
| plugin-docs-writer | Agent | Generate plugin documentation |
| context-gathering | Agent | Execute research tasks |
| subagent-contract | Skill | Enforce role boundaries and signaling |
| agent-creator | Skill | Create agents with proper format |
Gap Analysis: No automated error recovery, rollback, or retry mechanisms.
Stage 5: Verification ✅ STRONGEST
flowchart TB
subgraph SELF["Self-Assessment"]
SA1["verify skill<br/>Pre-completion checklist"]
SA2["/am-i-complete<br/>Completion readiness"]
SA3["/how-confident<br/>Confidence assessment"]
end
subgraph AUDIT["Audit Functions"]
AU1["audit skill<br/>Hallucination detection"]
AU2["/audit command<br/>Verify claims"]
end
subgraph REVIEW["Code/Asset Review"]
R1["code-review agent<br/>Security & quality"]
R2["plugin-assessor agent<br/>Plugin validation"]
R3["doc-drift-auditor agent<br/>Doc accuracy"]
end
subgraph GAP["Gaps"]
G1["❌ Performance verification"]
G2["❌ Benchmark testing"]
end
SA1 --> AU1
AU1 --> R1
R1 --> SA2
Coverage Details:
| Asset | Type | Function |
|---|---|---|
| verify | Skill | Rigorous self-assessment checklist |
| audit | Skill | Detect hallucinations and assumptions |
| /am-i-complete | Command | Check completion readiness |
| /verify | Command | Execute verification checklist |
| /audit | Command | Trigger hallucination audit |
| /how-confident | Command | Self-assess confidence level |
| code-review | Agent | Security, bugs, code quality |
| plugin-assessor | Agent | Plugin structure validation |
| doc-drift-auditor | Agent | Documentation accuracy |
Gap Analysis: No performance benchmarking or automated performance verification.
Stage 6: Output Delivery ⚠️ PARTIAL
flowchart TB
subgraph COMMIT["Git Operations"]
C1["git-commit-helper skill<br/>Commit message generation"]
end
subgraph SIGNAL["Completion Signaling"]
S1["DONE status<br/>Task complete"]
S2["BLOCKED status<br/>Cannot proceed"]
end
subgraph DOCS["Documentation"]
D1["plugin-docs-writer agent<br/>Generate README"]
D2["logging agent<br/>Consolidate work logs"]
end
subgraph GAP["Gaps"]
G1["❌ PR workflow"]
G2["❌ Changelog generation"]
G3["❌ Release notes"]
end
C1 --> S1
D1 --> S1
D2 --> S1
Coverage Details:
| Asset | Type | Function |
|---|---|---|
| git-commit-helper | Skill | Generate descriptive commit messages |
| subagent-contract (DONE/BLOCKED) | Skill | Signal completion status |
| plugin-docs-writer | Agent | Generate plugin documentation |
| logging | Agent | Consolidate work logs |
Gap Analysis: No PR creation workflow, no automated changelog, no release note generation.
Complete Asset-to-Stage Mapping
┌─────────────────────┬───────┬─────────┬──────────┬───────────┬──────────────┬────────┐
│ Asset │ Input │ Context │ Planning │ Execution │ Verification │ Output │
├─────────────────────┼───────┼─────────┼──────────┼───────────┼──────────────┼────────┤
│ SKILLS │ │ │ │ │ │ │
│ rt-ica │ │ ● │ ● │ │ │ │
│ delegate │ │ │ ● │ │ │ │
│ verify │ │ │ │ │ ● │ │
│ audit │ │ │ │ │ ● │ │
│ scientific-thinking │ │ │ ● │ ● │ │ │
│ agent-creator │ │ │ │ ● │ │ │
│ subagent-contract │ │ │ │ ● │ │ ● │
│ git-commit-helper │ │ │ │ │ │ ● │
│ *-reference-2026 x4 │ │ ● │ │ │ │ │
├─────────────────────┼───────┼─────────┼──────────┼───────────┼──────────────┼────────┤
│ AGENTS │ │ │ │ │ │ │
│ context-gathering │ │ ● │ │ ● │ │ │
│ context-refinement │ │ ● │ │ │ │ │
│ code-review │ │ │ │ │ ● │ │
│ plugin-assessor │ │ │ │ │ ● │ │
│ plugin-docs-writer │ │ │ │ ● │ │ ● │
│ skill-refactorer │ │ │ │ ● │ │ │
│ doc-drift-auditor │ │ │ │ │ ● │ │
│ context-optimizer │ │ │ │ ● │ │ │
│ subagent-refactorer │ │ │ │ ● │ │ │
│ subagent-generator │ │ │ │ ● │ │ │
│ logging │ │ │ │ │ │ ● │
├─────────────────────┼───────┼─────────┼──────────┼───────────┼──────────────┼────────┤
│ COMMANDS │ │ │ │ │ │ │
│ /sessions │ ● │ │ │ │ │ │
│ /am-i-complete │ │ │ │ │ ● │ │
│ /verify │ │ │ │ │ ● │ │
│ /audit │ │ │ │ │ ● │ │
│ /how-to-delegate │ │ │ ● │ │ │ │
│ /think │ │ │ ● │ │ │ │
│ /step-back │ │ │ ● │ │ │ │
│ /how-confident │ │ │ │ │ ● │ │
│ /rt-ica │ │ ● │ ● │ │ │ │
│ /delegate │ │ │ ● │ │ │ │
│ /scientific-thinking│ │ │ ● │ │ │ │
├─────────────────────┼───────┼─────────┼──────────┼───────────┼──────────────┼────────┤
│ HOOKS │ │ │ │ │ │ │
│ session-start-rtica │ ● │ │ │ │ │ │
├─────────────────────┼───────┼─────────┼──────────┼───────────┼──────────────┼────────┤
│ COVERAGE SCORE │ 2/3 │ 7/3 │ 8/3 │ 8/3 │ 9/3 │ 4/3 │
│ STATUS │ ⚠️ │ ✅ │ ✅ │ ✅ │ ✅ │ ⚠️ │
└─────────────────────┴───────┴─────────┴──────────┴───────────┴──────────────┴────────┘
Legend: ● = Asset covers this stage
Coverage threshold: ≥3 assets = ✅ COVERED, 1-2 assets = ⚠️ PARTIAL, 0 = ❌ GAP
Navigation
- Next: Asset Decision Tree - How to choose the right asset type
- Back to: Index