
Capability Discovery
- 6 installs
- 4 repo stars
- Updated June 18, 2026
- doubleslashse/claude-marketplace
Discover installed plugins at runtime and map them to abstract capability categories to route tasks to agents without hardcoding plugin names.
About
Enumerates installed plugins and maps them to capability categories with graceful fallback and explicit routing announcements. A workflow uses it to delegate tasks to the right agent dynamically.
- Maps capabilities, not specific plugin names
- Announces delegation decisions with graceful fallback
Capability Discovery by the numbers
- 6 all-time installs (skills.sh)
- Ranked #12,825 of 16,546 AI & Agent Building skills by installs in the Skillselion catalog
- Data as of Jul 29, 2026 (Skillselion catalog sync)
npx skills add https://github.com/doubleslashse/claude-marketplace --skill capability-discoveryAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 6 |
|---|---|
| repo stars | ★ 4 |
| Last updated | June 18, 2026 |
| Repository | doubleslashse/claude-marketplace ↗ |
What it does
Discover installed plugins at runtime and map them to abstract capability categories to route tasks to agents without hardcoding plugin names.
Files
Capability Discovery Skill
This skill enables dynamic discovery and mapping of installed plugins to abstract capability categories, allowing the workflow to route tasks to appropriate agents without hardcoding plugin names. It includes explicit delegation announcements so users understand routing decisions.
Core Principles
1. Technology Agnostic: Map capabilities, not specific plugins 2. Dynamic Discovery: Scan at runtime, not compile time 3. Graceful Fallback: Use internal default agents when no plugin matches 4. Project-Aware Routing: Prefer plugins matching detected project type 5. Transparent Delegation: Always announce why a particular agent was chosen
Capability Categories
| Capability | Purpose | Phase | Default Agent |
|---|---|---|---|
requirements-gathering | Structured requirements elicitation | DISCUSS | defaults/interviewer |
brainstorming | Ideation and option exploration | DISCUSS | defaults/interviewer |
codebase-analysis | Understanding existing code | DISCUSS, PLAN | defaults/researcher |
tdd-implementation | Test-driven development | EXECUTE | defaults/executor |
code-implementation | General code writing | EXECUTE | defaults/executor |
infrastructure | DevOps/infra tasks | EXECUTE | defaults/executor |
code-review | Quality validation | VERIFY | validator |
requirements-validation | Requirements compliance | VERIFY | validator |
Discovery Process
Step 1: Enumerate Installed Plugins
1. Locate Plugins/ directory
2. List all subdirectories (each is a plugin)
3. For each plugin:
a. Read .claude-plugin/plugin.json for metadata
b. Scan agents/*.md for agent definitions
c. Scan commands/*.md for command definitions
4. Build plugin registryStep 2: Extract Descriptions
For each agent/command file, extract the description from YAML frontmatter:
---
name: implementer
description: TDD implementation specialist for .NET. Use to write minimal code that makes tests pass.
model: opus
tools: [Read, Write, Edit, Bash, Grep, Glob]
---Step 3: Map to Capabilities (Keyword Scoring)
Apply keyword matching to descriptions:
For each plugin agent/command:
1. Tokenize description into words (lowercase)
2. Match against capability keywords
3. Calculate score:
- Exact keyword match: +10 points
- Partial match (word contains keyword): +3 points
- Keyword in first sentence: +5 bonus
- Multiple keywords from same category: +2 each after first
4. Assign to highest-scoring capability
5. Record match confidence (High: 25+, Medium: 15-24, Low: <15)Step 4: Detect Project Type
Scan project root for technology indicators:
| File Pattern | Project Type | Keywords |
|---|---|---|
*.csproj, *.sln | dotnet | .NET, dotnet, C#, csharp |
package.json, *.js, *.ts | node | Node.js, JavaScript, TypeScript |
*.py, requirements.txt | python | Python, pip, pytest |
go.mod, *.go | go | Go, golang |
Cargo.toml, *.rs | rust | Rust, Cargo |
pom.xml, build.gradle | java | Java, Maven, Gradle |
Step 5: Cache Results in FLOW.md
Store capability mappings in the Capabilities Cache section of FLOW.md.
Delegation Priority System
When routing a task to an agent, follow this priority order:
Priority 1: Project-Type Match
If project is detected (e.g., dotnet), prefer plugins with matching technology keywords in their description.
Detected: dotnet project
Capability needed: tdd-implementation
Matches:
- dotnet-tdd:implementer (description: "TDD for .NET") → Score: 35 + project bonus
- node-tdd:implementer (description: "TDD for Node.js") → Score: 30
Winner: dotnet-tdd:implementer (project type match)Priority 2: Exact Keyword Match
Prefer agents with exact keyword matches over partial matches.
Priority 3: Agent Over Command
When both an agent and a command match a capability, prefer the agent (more capable).
Priority 4: Internal Default
When no plugin matches the capability, use the appropriate default agent.
Delegation Announcement Pattern
CRITICAL: Always announce delegation decisions to the user with reasoning.
When Plugin Found
**Delegating [capability]** → [plugin-name]:[agent-name]
Matched via keyword scoring:
- Keywords: [matched keywords]
- Score: [score] (High confidence)
- Project type: [type] (matched)
Spawning agent...When Using Default
**Using built-in agent** for [capability] → flow-workflow:defaults/[agent]
No installed plugin matched this capability.
- Searched: [N] plugins
- Keywords tried: [keywords]
Consider installing: [suggested plugin type]Quick Delegation (For Status/Logs)
→ [capability]: [plugin:agent] (keyword match)
→ [capability]: defaults/[agent] (no plugin match)Routing Logic
When Phase Needs Capability
1. Determine capability needed for current task
2. Look up FLOW.md capability cache
3. If cache stale (>24h) or missing, re-scan plugins
4. If multiple matches for capability:
a. Filter by project type preference
b. Prefer exact keyword matches
c. Prefer agents over commands
5. If single match:
a. Announce delegation with reasoning
b. Return agent reference
6. If no match:
a. Announce using default
b. Return default agent referenceCapability → Phase Mapping
| Phase | Primary Capability | Secondary Capabilities |
|---|---|---|
| DISCUSS | requirements-gathering | brainstorming, codebase-analysis |
| PLAN | (internal coordinator) | codebase-analysis |
| EXECUTE | code-implementation | tdd-implementation, infrastructure |
| VERIFY | code-review | requirements-validation |
Default Agents
When no plugin matches a capability:
| Capability | Default Agent | Location |
|---|---|---|
| requirements-gathering | interviewer | flow-workflow:defaults/interviewer |
| brainstorming | interviewer | flow-workflow:defaults/interviewer |
| codebase-analysis | researcher | flow-workflow:defaults/researcher |
| tdd-implementation | executor | flow-workflow:defaults/executor |
| code-implementation | executor | flow-workflow:defaults/executor |
| infrastructure | executor | flow-workflow:defaults/executor |
| code-review | validator | flow-workflow:validator |
| requirements-validation | validator | flow-workflow:validator |
Re-Discovery Triggers
Re-scan plugins when:
1. /flow-workflow:start is run on uninitialized project 2. FLOW.md capability cache is older than 24 hours 3. Requested capability returns no match (try fresh scan) 4. User explicitly requests refresh via /flow-workflow:status --refresh
Capability Gap Logging
When no plugin matches a capability, log in FLOW.md:
## Capability Warnings
**[TIMESTAMP]**: No plugin found for capability 'infrastructure'
- Searched: 5 plugins
- Keywords tried: infra, devops, pipeline, deploy
- Using: flow-workflow:defaults/executor
- Suggestion: Consider installing a DevOps pluginIntegration Points
- State Management: Cache capability map in FLOW.md
- Coordinator Agent: Performs discovery on initialization, uses for routing
- Smart Continuation: Uses cached capabilities for agent selection
See categories.md for detailed capability category definitions.
Capability Category Definitions
Overview
Capability categories provide an abstraction layer between workflow phases and specific plugin implementations. This enables technology-agnostic orchestration with transparent delegation.
Category Definitions
requirements-gathering
Purpose: Structured elicitation of project requirements from stakeholders
When Used: DISCUSS phase, when understanding what to build
Keywords: requirements, interview, elicitation, stakeholder, gathering, discover, needs
Expected Behaviors:
- Conduct structured interviews
- Ask clarifying questions
- Document functional requirements
- Document non-functional requirements
- Identify constraints and assumptions
Default Agent: flow-workflow:defaults/interviewer
---
codebase-analysis
Purpose: Understanding existing code, patterns, and architecture
When Used: DISCUSS phase (for brownfield projects), PLAN phase (for implementation planning)
Keywords: codebase, analyze, reverse-engineer, patterns, architecture, existing, legacy
Expected Behaviors:
- Scan project structure
- Identify design patterns
- Detect technologies in use
- Find integration points
- Document existing behavior
Default Agent: flow-workflow:defaults/researcher
---
brainstorming
Purpose: Creative exploration of ideas and options
When Used: DISCUSS phase, when exploring possible approaches
Keywords: brainstorm, ideation, workshop, creative, explore, options, ideas
Expected Behaviors:
- Facilitate idea generation
- Divergent thinking exercises
- Option comparison
- Trade-off analysis
- Convergent selection
Default Agent: flow-workflow:defaults/interviewer (with brainstorming prompts)
---
tdd-implementation
Purpose: Test-driven development with RED-GREEN-REFACTOR cycle
When Used: EXECUTE phase, when implementing with tests first
Keywords: tdd, test-driven, red-green, refactor, tests first, failing test
Expected Behaviors:
- Write failing tests first (RED)
- Implement minimal code to pass (GREEN)
- Refactor for quality (REFACTOR)
- Maintain test coverage
- Follow clean code principles
Default Agent: flow-workflow:defaults/executor (with TDD guidance)
---
code-implementation
Purpose: General code writing and modification
When Used: EXECUTE phase, for any coding task
Keywords: implement, developer, code, write, create, build, feature
Expected Behaviors:
- Write production code
- Follow project conventions
- Handle errors appropriately
- Integrate with existing code
- Document as needed
Default Agent: flow-workflow:defaults/executor
---
infrastructure
Purpose: DevOps, CI/CD, and infrastructure tasks
When Used: EXECUTE phase, for deployment and operations
Keywords: infra, devops, pipeline, deploy, docker, kubernetes, ci, cd, terraform, azure
Expected Behaviors:
- Configure build pipelines
- Write deployment scripts
- Manage infrastructure as code
- Set up environments
- Configure monitoring
Default Agent: flow-workflow:defaults/executor (with infrastructure guidance)
---
code-review
Purpose: Quality validation of implemented code
When Used: VERIFY phase, after implementation
Keywords: review, quality, clean code, solid, dry, kiss, standards
Expected Behaviors:
- Review code for quality
- Check adherence to standards
- Identify potential issues
- Suggest improvements
- Validate best practices
Default Agent: flow-workflow:validator
---
requirements-validation
Purpose: Verify implementation meets requirements
When Used: VERIFY phase, for acceptance testing
Keywords: validate, verification, compliance, acceptance, requirements, uat
Expected Behaviors:
- Check requirements coverage
- Validate acceptance criteria
- Identify gaps
- Conduct UAT
- Document compliance
Default Agent: flow-workflow:validator
---
Keyword Matching Algorithm
Scoring Rules
1. Exact keyword match: +10 points
2. Partial match (word contains keyword): +3 points
3. Keyword in first sentence: +5 bonus
4. Multiple keywords from same category: +2 each after first
5. Project type match: +10 bonusConfidence Levels
| Score | Confidence | Meaning |
|---|---|---|
| 25+ | High | Strong match, safe to delegate |
| 15-24 | Medium | Good match, may need verification |
| <15 | Low | Weak match, consider default |
Matching Example
Description: "TDD implementation specialist for .NET. Write minimal code that makes tests pass."
Keyword matches:
- "TDD" → tdd-implementation (+10)
- "implementation" → code-implementation (+10)
- "tests" → tdd-implementation (+10)
- ".NET" → (project type indicator)
Scores:
- tdd-implementation: 20 + 5 (first sentence) = 25 (High)
- code-implementation: 10 (Low)
Winner: tdd-implementation---
Project Type Routing
When Multiple Capabilities Match
If multiple plugins match a capability, filter by project type:
Capability: tdd-implementation
Matches:
- dotnet-tdd:implementer (description contains ".NET")
- node-tdd:implementer (description contains "Node.js")
Detected project type: dotnet
Selected: dotnet-tdd:implementer (+10 project bonus)Project Type Keywords
| Project Type | Description Keywords |
|---|---|
| dotnet | .NET, dotnet, C#, csharp, NuGet |
| node | Node.js, node, JavaScript, TypeScript, npm, yarn |
| python | Python, pip, pytest, Django, Flask |
| go | Go, golang, Gin, Echo |
| rust | Rust, Cargo, crate |
| java | Java, Maven, Gradle, Spring |
---
Default Agent Summary
| Capability | Default Agent | Why |
|---|---|---|
| requirements-gathering | defaults/interviewer | General interview skills |
| brainstorming | defaults/interviewer | Facilitation skills overlap |
| codebase-analysis | defaults/researcher | General research skills |
| tdd-implementation | defaults/executor | TDD guidance built in |
| code-implementation | defaults/executor | General coding skills |
| infrastructure | defaults/executor | Infra guidance built in |
| code-review | validator | Review is built-in function |
| requirements-validation | validator | UAT is built-in function |
---
Capability Gap Handling
When No Plugin Matches
1. Log warning to FLOW.md with timestamp 2. Announce to user: "Using built-in agent (no plugin matched)" 3. Use appropriate default agent 4. Suggest plugin installation if capability is important
Suggested Plugins by Capability
| Missing Capability | Suggested Plugin Type |
|---|---|
| requirements-gathering | business-analyst |
| brainstorming | workshop-facilitator |
| codebase-analysis | business-analyst |
| tdd-implementation | dotnet-tdd, node-tdd |
| code-implementation | developer plugin for project type |
| infrastructure | devops-azure-infrastructure, infra-plugin |
| code-review | tdd plugin with reviewer |
| requirements-validation | business-analyst |
---
Delegation Announcement Templates
Plugin Match
**Delegating requirements-gathering** → business-analyst:stakeholder-interviewer
Matched via keyword scoring:
- Keywords: requirements, interview, stakeholder
- Score: 30 (High confidence)
- Project type: dotnet (neutral)Default Agent
**Using built-in agent** for infrastructure → flow-workflow:defaults/executor
No installed plugin matched this capability.
- Searched: 5 plugins
- Keywords tried: infra, devops, pipeline, deploy
Consider installing: devops-azure-infrastructure or similarShort Form (For Logs)
→ requirements-gathering: business-analyst:stakeholder-interviewer (keyword match)
→ infrastructure: defaults/executor (no plugin match)