Task Analyzer
Provides metacognitive task analysis and skill selection guidance.
Skills Index
See skills-index.yaml for available skills metadata.
Task Analysis Process
1. Understand Task Essence
Identify the fundamental purpose beyond surface-level work:
| Surface Work |
Fundamental Purpose |
| "Fix this bug" |
Problem solving, root cause analysis |
| "Implement this feature" |
Feature addition, value delivery |
| "Refactor this code" |
Quality improvement, maintainability |
| "Update this file" |
Change management, consistency |
Key Questions:
- What problem are we really solving?
- What is the expected outcome?
- What could go wrong if we approach this superficially?
2. Estimate Structural Scale
Classify decision burden from the intended outcomes and responsibility boundaries. File count is supporting evidence only.
| Scale |
Decision burden |
| Small |
One coherent outcome, one evident repository-supported implementation within one responsibility boundary, and no unresolved durable choice |
| Medium |
One coherent outcome that coordinates a boundary or contains a potentially durable choice |
| Large |
Multiple independently valuable outcomes that require separate design decisions |
A cross-layer implementation can remain Medium when it serves one coherent outcome. A decision point passing both documentation-criteria ADR filters raises the scale to Medium at minimum. Record the evidence that established the outcome and boundary classification in scaleRationale.
Scale affects skill priority:
- Larger scale → process/documentation skills more important
- Smaller scale → implementation skills more focused
3. Identify Task Type
| Type |
Characteristics |
Key Skills |
| implementation |
New code or user-visible behavior |
coding-standards, typescript-testing |
| fix |
Defect or regression resolution |
coding-standards, typescript-testing |
| refactoring |
Behavior-preserving structure improvement |
coding-standards, implementation-approach |
| design |
Architecture or contract decisions |
documentation-criteria, implementation-approach |
| quality |
Testing, review, verification |
typescript-testing, integration-e2e-testing |
| documentation |
PRD, ADR, Design Doc, UI Spec, plan, or instruction content |
documentation-criteria |
| investigation |
Evidence gathering without implementation |
project-context plus the domain skill selected from the index |
| migration |
Data, schema, API, dependency, or runtime transition |
implementation-approach, documentation-criteria |
| operations |
Environment, deployment, or runtime operation |
technical-spec plus the domain skill selected from the index |
| security |
Security design or review |
coding-standards plus the implementation-domain skill |
| skill |
Skill creation, prompt-quality review, or skill metadata change |
skill-optimization, llm-friendly-context |
When multiple types apply, return the primary type that owns the requested outcome and list the remaining values in secondaryTypes.
4. Tag-Based Skill Matching
Extract relevant tags from task description and match against skills-index.yaml:
Task: "Implement user authentication with tests"
Extracted tags: [implementation, testing, security]
Matched skills:
- coding-standards (implementation, security)
- typescript-testing (testing)
- typescript-rules (implementation)
5. Implicit Relationships
Consider hidden dependencies:
| Task Involves |
Also Include |
| Error handling |
debugging, testing |
| New features |
design, implementation, documentation |
| Performance |
profiling, optimization, testing |
| Frontend |
typescript-rules, typescript-testing |
| API/Integration |
integration-e2e-testing |
Output Format
Return structured analysis with skill metadata from skills-index.yaml:
taskAnalysis:
essence: <string> # Fundamental purpose identified
type: <implementation|fix|refactoring|design|quality|documentation|investigation|migration|operations|security|skill>
secondaryTypes: [<task-type>, ...]
scale: <small|medium|large>
estimatedFiles: <number or unknown> # Supporting evidence only
scaleRationale:
decidingAxis: <outcomes|responsibility-boundaries|durable-choice>
evidence: <string>
tags: [<string>, ...] # Extracted from task description
selectedSkills:
- skill: <skill-name> # From skills-index.yaml
priority: <high|medium|low>
reason: <string> # Why this skill was selected
# Pass through metadata from skills-index.yaml
tags: [...]
typical-use: <string>
size: <small|medium|large>
sections: [...] # All sections from yaml, unfiltered
Note: Section selection (choosing which sections are relevant) is done separately after reading the actual SKILL.md files.
Process Gates
- Intent gate: Proceed to scale estimation when
essence, primary type, and any secondaryTypes are recorded. If the requested outcome is ambiguous, record the exact outcome decision required.
- Scale gate: Proceed to skill matching when the outcome and responsibility-boundary evidence is sufficient for Structural Scale and
scaleRationale names the deciding axis.
- Selection gate: Finalize when every selected skill exists in
skills-index.yaml, has a reason tied to the task, and its metadata is copied without invention.
When an unknown can change the outcome boundary, ADR qualification, or required workflow, request the exact repository evidence or user decision needed. An unknown file count alone does not block Structural Scale judgment.
Skill Selection Priority
- Essential - Directly related to task type
- Quality - Testing and quality assurance
- Process - Workflow and documentation
- Supplementary - Additional constraints or evidence directly tied to the task
Metacognitive Question Design
Generate only questions whose answers can change intent classification, scale, selected skills, a hard constraint, or verification. Return no question when repository evidence already resolves those decisions. For every question, record the decision it controls.
| Task Type |
Question Focus |
| Implementation |
Design validity, edge cases, performance |
| Fix |
Root cause (5 Whys), impact scope, regression testing |
| Refactoring |
Current problems, target state, phased plan |
| Design |
Requirement clarity, trade-offs |
| Documentation |
Audience, source of truth, approval/consumer contract |
| Investigation |
Claim to resolve, evidence boundary, stopping condition |
| Migration |
Compatibility window, data/contract transition, rollback |
| Operations |
Target environment, authorization boundary, recovery evidence |
| Security |
Trust boundary, protected asset, threat source, risk-acceptance authority |
| Skill |
Triggering intent, standalone context, output consumer |
Warning Patterns
Detect and flag these patterns:
| Pattern |
Warning |
Mitigation |
| One step contains multiple independently verifiable outcomes |
Transition and rollback risk |
Split at observable verification boundaries |
| A behavior change has no test or named runnable verification |
Regression evidence is missing |
Add the cheapest check that observes the changed contract |
| A proposed fix has no observed causal link to the failure |
Root cause remains inferred |
Record reproduction evidence and the first causal boundary before selecting the fix |
| Medium/Large implementation lacks its scale-required planning artifact |
Scope and dependency contract is missing |
Create the required artifact before implementation routing |
1---2name: task-analyzer3description: Classifies task intent, change risk, and execution scale, then selects skills from the project skills index. Use when starting work, routing a task, estimating scope, or selecting skills.4---56# Task Analyzer78Provides metacognitive task analysis and skill selection guidance.910## Skills Index1112See **[skills-index.yaml](references/skills-index.yaml)** for available skills metadata.1314## Task Analysis Process1516### 1. Understand Task Essence1718Identify the fundamental purpose beyond surface-level work:1920| Surface Work | Fundamental Purpose |21|--------------|---------------------|22| "Fix this bug" | Problem solving, root cause analysis |23| "Implement this feature" | Feature addition, value delivery |24| "Refactor this code" | Quality improvement, maintainability |25| "Update this file" | Change management, consistency |2627**Key Questions:**28- What problem are we really solving?29- What is the expected outcome?30- What could go wrong if we approach this superficially?3132### 2. Estimate Structural Scale3334Classify decision burden from the intended outcomes and responsibility boundaries. File count is supporting evidence only.3536| Scale | Decision burden |37|-------|-----------------|38| Small | One coherent outcome, one evident repository-supported implementation within one responsibility boundary, and no unresolved durable choice |39| Medium | One coherent outcome that coordinates a boundary or contains a potentially durable choice |40| Large | Multiple independently valuable outcomes that require separate design decisions |4142A cross-layer implementation can remain Medium when it serves one coherent outcome. A decision point passing both documentation-criteria ADR filters raises the scale to Medium at minimum. Record the evidence that established the outcome and boundary classification in `scaleRationale`.4344**Scale affects skill priority:**45- Larger scale → process/documentation skills more important46- Smaller scale → implementation skills more focused4748### 3. Identify Task Type4950| Type | Characteristics | Key Skills |51|------|-----------------|------------|52| implementation | New code or user-visible behavior | coding-standards, typescript-testing |53| fix | Defect or regression resolution | coding-standards, typescript-testing |54| refactoring | Behavior-preserving structure improvement | coding-standards, implementation-approach |55| design | Architecture or contract decisions | documentation-criteria, implementation-approach |56| quality | Testing, review, verification | typescript-testing, integration-e2e-testing |57| documentation | PRD, ADR, Design Doc, UI Spec, plan, or instruction content | documentation-criteria |58| investigation | Evidence gathering without implementation | project-context plus the domain skill selected from the index |59| migration | Data, schema, API, dependency, or runtime transition | implementation-approach, documentation-criteria |60| operations | Environment, deployment, or runtime operation | technical-spec plus the domain skill selected from the index |61| security | Security design or review | coding-standards plus the implementation-domain skill |62| skill | Skill creation, prompt-quality review, or skill metadata change | skill-optimization, llm-friendly-context |6364When multiple types apply, return the primary type that owns the requested outcome and list the remaining values in `secondaryTypes`.6566### 4. Tag-Based Skill Matching6768Extract relevant tags from task description and match against skills-index.yaml:6970```yaml71Task: "Implement user authentication with tests"72Extracted tags: [implementation, testing, security]73Matched skills:74 - coding-standards (implementation, security)75 - typescript-testing (testing)76 - typescript-rules (implementation)77```7879### 5. Implicit Relationships8081Consider hidden dependencies:8283| Task Involves | Also Include |84|---------------|--------------|85| Error handling | debugging, testing |86| New features | design, implementation, documentation |87| Performance | profiling, optimization, testing |88| Frontend | typescript-rules, typescript-testing |89| API/Integration | integration-e2e-testing |9091## Output Format9293Return structured analysis with skill metadata from skills-index.yaml:9495```yaml96taskAnalysis:97 essence: <string> # Fundamental purpose identified98 type: <implementation|fix|refactoring|design|quality|documentation|investigation|migration|operations|security|skill>99 secondaryTypes: [<task-type>, ...]100 scale: <small|medium|large>101 estimatedFiles: <number or unknown> # Supporting evidence only102 scaleRationale:103 decidingAxis: <outcomes|responsibility-boundaries|durable-choice>104 evidence: <string>105 tags: [<string>, ...] # Extracted from task description106107selectedSkills:108 - skill: <skill-name> # From skills-index.yaml109 priority: <high|medium|low>110 reason: <string> # Why this skill was selected111 # Pass through metadata from skills-index.yaml112 tags: [...]113 typical-use: <string>114 size: <small|medium|large>115 sections: [...] # All sections from yaml, unfiltered116```117118**Note**: Section selection (choosing which sections are relevant) is done separately after reading the actual SKILL.md files.119120## Process Gates1211221. **Intent gate**: Proceed to scale estimation when `essence`, primary `type`, and any `secondaryTypes` are recorded. If the requested outcome is ambiguous, record the exact outcome decision required.1232. **Scale gate**: Proceed to skill matching when the outcome and responsibility-boundary evidence is sufficient for Structural Scale and `scaleRationale` names the deciding axis.1243. **Selection gate**: Finalize when every selected skill exists in `skills-index.yaml`, has a reason tied to the task, and its metadata is copied without invention.125126When an unknown can change the outcome boundary, ADR qualification, or required workflow, request the exact repository evidence or user decision needed. An unknown file count alone does not block Structural Scale judgment.127128## Skill Selection Priority1291301. **Essential** - Directly related to task type1312. **Quality** - Testing and quality assurance1323. **Process** - Workflow and documentation1334. **Supplementary** - Additional constraints or evidence directly tied to the task134135## Metacognitive Question Design136137Generate only questions whose answers can change intent classification, scale, selected skills, a hard constraint, or verification. Return no question when repository evidence already resolves those decisions. For every question, record the decision it controls.138139| Task Type | Question Focus |140|-----------|----------------|141| Implementation | Design validity, edge cases, performance |142| Fix | Root cause (5 Whys), impact scope, regression testing |143| Refactoring | Current problems, target state, phased plan |144| Design | Requirement clarity, trade-offs |145| Documentation | Audience, source of truth, approval/consumer contract |146| Investigation | Claim to resolve, evidence boundary, stopping condition |147| Migration | Compatibility window, data/contract transition, rollback |148| Operations | Target environment, authorization boundary, recovery evidence |149| Security | Trust boundary, protected asset, threat source, risk-acceptance authority |150| Skill | Triggering intent, standalone context, output consumer |151152## Warning Patterns153154Detect and flag these patterns:155156| Pattern | Warning | Mitigation |157|---------|---------|------------|158| One step contains multiple independently verifiable outcomes | Transition and rollback risk | Split at observable verification boundaries |159| A behavior change has no test or named runnable verification | Regression evidence is missing | Add the cheapest check that observes the changed contract |160| A proposed fix has no observed causal link to the failure | Root cause remains inferred | Record reproduction evidence and the first causal boundary before selecting the fix |161| Medium/Large implementation lacks its scale-required planning artifact | Scope and dependency contract is missing | Create the required artifact before implementation routing |
shinpr