Skills Reference — Smart SDLC
Every skill in the framework, with its type, persona access, activation method, and what it produces.
Skill types:
- Agent — activates a full persona with menu-driven interaction
- Workflow — drives a specific multi-step task
- Utility — cross-cutting tools available to all personas
- Integration — connects to external systems (JIRA, Confluence, GitHub, etc.)
Activation methods (GitHub Copilot):
@sml-<name> — agent picker (for agent-type skills)/sml-<name> — slash command#file:_superml/skills/<path>/SKILL.md — direct file attachment
Phase 0 — Relearn (Brownfield Onboarding)
All skills in this phase are unrestricted — any persona can use them.
| Skill |
Command |
Type |
What It Does |
Output |
| agent-scout |
@sml-agent-scout |
Agent |
Code Archaeologist persona. Onboarding to unfamiliar codebases, reverse-engineering docs. |
Interactive menu |
| relearn-codebase |
/sml-relearn-codebase |
Workflow |
Full 5-step reverse-engineering: scan structure, architecture, data/API, patterns, produce docs. |
docs/legacy/*.md or docs/*.md |
| generate-readme |
/sml-generate-readme |
Workflow |
Generate or refresh README.md from the actual codebase — setup, env, scripts, project structure. |
README.md |
| generate-api-docs |
/sml-generate-api-docs |
Workflow |
Generate API reference documentation (markdown or OpenAPI YAML) from route files and schemas. |
docs/api-reference.md or openapi.yaml |
| reverse-adr |
/sml-reverse-adr |
Workflow |
Reverse-engineer Architecture Decision Records from code evidence — document WHY things are this way. |
docs/decisions/ADR-*.md |
relearn-codebase — Step Detail
| Step |
What Scout Does |
| 01 — Scan |
File tree, tech stack detection, entry points, test coverage indicators |
| 02 — Architecture |
Component diagram, layer map, external service dependencies |
| 03 — Data & API |
Data models, DB schemas, API routes and contracts |
| 04 — Patterns |
Coding conventions, error handling patterns, naming rules, anti-patterns found |
| 05 — Document |
Consolidates all above into structured markdown suite |
Phase 1 — Analysis
| Skill |
Command |
Type |
Persona |
What It Does |
Output |
| agent-analyst |
@sml-agent-analyst |
Agent |
product |
Business Analyst persona. Requirements elicitation, codebase analysis, competitive research. |
Interactive menu |
| document-project |
/sml-document-project |
Workflow |
All |
Analyse an existing codebase and produce structured project documentation for human and AI consumption. |
docs/project-overview.md |
| product-brief |
/sml-product-brief |
Workflow |
product |
Guided discovery to create a product brief: problem, users, metrics, scope. |
_superml-output/planning/product-brief.md |
Phase 2 — Planning
| Skill |
Command |
Type |
Persona |
What It Does |
Output |
| agent-pm |
@sml-agent-pm |
Agent |
product |
Product Manager persona. PRD creation, requirements discovery, stakeholder alignment. |
Interactive menu |
| create-prd |
/sml-create-prd |
Workflow |
product |
3-step guided PRD: goals & users, features & flows, non-functional requirements. |
_superml-output/planning/prd-<name>.md |
| edit-prd |
/sml-edit-prd |
Workflow |
product |
Refine an existing PRD section by section without rewriting the whole document. |
Updated prd-*.md |
| validate-prd |
/sml-validate-prd |
Workflow |
product |
Validate PRD for completeness, consistency, and implementation readiness. Flags gaps. |
Validation report inline |
create-prd — Step Detail
| Step |
What Gets Written |
| Goals & Users |
Objective, user segments, user journeys, success criteria |
| Features & Flows |
Feature list, user flows per feature, acceptance criteria |
| Non-Functional |
Performance, security, compatibility, constraints, timeline |
Phase 3 — Solutioning
| Skill |
Command |
Type |
Persona |
What It Does |
Output |
| agent-architect |
@sml-agent-architect |
Agent |
architect |
System Architect persona. Technical architecture, system design, technology decisions. |
Interactive menu |
| create-architecture |
/sml-create-architecture |
Workflow |
architect |
Full architecture document: components, data, API, ADRs, non-functional, risks. |
_superml-output/planning/architecture-*.md |
| create-epics-stories |
/sml-create-epics-stories |
Workflow |
architect, team_lead |
Break PRD + architecture into epics and detailed user stories with acceptance criteria. |
_superml-output/planning/epics-stories.md |
| generate-context |
/sml-generate-context |
Workflow |
architect |
Generate a compact project-context.md optimised for AI context loading across sessions. |
project-context.md |
create-architecture — Section Detail
| Section |
Content |
| System Context |
What the system does, who uses it, external dependencies |
| Component Design |
Services / modules / layers, responsibilities, contracts |
| Data Design |
Entity model, schema, storage choices, migration plan |
| API Design |
Endpoints, request/response contracts, versioning, auth |
| Infrastructure |
Deployment topology, environments, CI/CD |
| ADRs |
Key technical decisions with rationale and rejected alternatives |
| Non-Functional |
Security, performance, observability, scalability |
| Implementation Notes |
Phasing suggestions, risks, open questions |
Phase 4 — Implementation
| Skill |
Command |
Type |
Persona |
What It Does |
Output |
| agent-developer |
@sml-agent-developer |
Agent |
developer |
Senior Developer persona. Story execution, TDD, code quality. |
Interactive menu |
| dev-story |
/sml-dev-story |
Workflow |
developer |
Implement a story using TDD (Red → Green → Refactor). |
Implemented code + commit |
| code-review |
/sml-code-review |
Workflow |
developer |
Structured code review against acceptance criteria, ADRs, and coding standards. |
Review report inline |
| create-story |
/sml-create-story |
Workflow |
developer, team_lead |
Write a single detailed story with full acceptance criteria and technical notes. |
Story doc or JIRA ticket |
| sprint-planning |
/sml-sprint-planning |
Workflow |
developer, team_lead |
Plan a sprint: select stories, estimate, assign, set sprint goal. |
_superml-output/implementation/sprint-<n>.md |
dev-story — Loop Detail
| Step |
What Happens |
| Load |
Read story, acceptance criteria, architecture context, coding standards |
| Red |
Write failing tests covering acceptance criteria |
| Green |
Write minimum implementation to pass tests |
| Refactor |
Improve structure, naming, readability while tests stay green |
| Review gate |
Self-checklist: AC ✔, code quality ✔, no console logs/dead code ✔ |
| Commit |
Semantic commit message referencing story ID |
| JIRA (if enabled) |
Transition ticket to In Review |
Phase 5 — Modernization
| Skill |
Command |
Type |
Persona |
What It Does |
Output |
| agent-sage |
@sml-agent-sage |
Agent |
modernization |
Modernization Lead persona. Legacy analysis, migration planning, knowledge extraction. |
Interactive menu |
| read-legacy-code |
/sml-read-legacy-code |
Workflow |
modernization |
Structured deep-read of legacy programs: entry points, control flow, business rules, data flows. |
docs/legacy/program-inventory.md, business-rules-raw.md |
| build-knowledge-graph |
/sml-build-knowledge-graph |
Workflow |
modernization |
Structure extracted rules into entity model, rule catalogue, process flows, decision tables. |
docs/legacy/knowledge-graph.md, process-flows/, decision-tables/ |
| validate-business-rules |
/sml-validate-business-rules |
Workflow |
modernization |
Walk stakeholders through extracted rules — confirm, dispute, or flag as incomplete. |
docs/legacy/business-rules-validated.md |
| define-target-architecture |
/sml-define-target-architecture |
Workflow |
modernization |
Assess legacy pain points, choose migration strategy, design target state, plan cut-over. |
_superml-output/planning/target-architecture.md |
| create-migration-epics |
/sml-create-migration-epics |
Workflow |
modernization |
Break migration plan into epics and stories with traceability back to business rules. |
_superml-output/planning/migration-epics.md |
Core Utility Skills
Available to all personas.
| Skill |
Command |
Type |
What It Does |
| help |
/sml-help |
Utility |
Context-aware guidance. Reads config.yml and persona.yml to tell you what to do next based on current project state. |
| brainstorming |
/sml-brainstorming |
Utility |
Facilitated brainstorming using diverge/converge technique. Useful before product brief or architecture. |
| elicitation |
/sml-elicitation |
Utility |
Advanced requirements elicitation: assumption surfacing, five-whys, stakeholder mapping, jobs-to-be-done. |
/sml-help — State Logic
/sml-help reads _superml/config.yml and _superml/persona.yml to detect which of three states the project is in:
| State |
Condition |
What Help Says |
| State A — No config |
Neither config.yml nor persona.yml exists |
Run init first |
| State B — Config exists, no persona |
config.yml exists but no persona.yml |
Run persona to set up your role |
| State C — Fully configured |
Both files exist |
Shows your persona’s next recommended action based on artifact completion flags |
Integration Skills
See Integrations for full setup and usage.
JIRA
| Skill |
Command |
What It Does |
| jira-connect |
/sml-jira-connect |
Set up JIRA connectivity — choose REST API (email + token) or MCP Server (Atlassian Remote / @sooperset/mcp-atlassian / custom URL) |
| jira-create-epic |
/sml-jira-create-epic |
Create an epic in JIRA from a Smart SDLC epics artifact |
| jira-create-story |
/sml-jira-create-story |
Create a JIRA story from a Smart SDLC story file |
| jira-sync |
/sml-jira-sync |
Sync all epics and stories from planning artifacts to JIRA |
| jira-conflict-detect |
/sml-jira-conflict-detect |
Check JIRA ticket status before branch creation to prevent work conflicts |
Confluence
| Skill |
Command |
What It Does |
| confluence-connect |
/sml-confluence-connect |
Set up Confluence connectivity — REST API or MCP Server; auto-detects if JIRA is already using Atlassian MCP and offers to reuse it |
| confluence-push-doc |
/sml-confluence-push-doc |
Push a planning artifact (PRD, architecture, etc.) to a Confluence page |
GitHub
| Skill |
Command |
What It Does |
| github-connect |
/sml-github-connect |
Verify gh CLI authentication and repo access |
| github-create-branch |
/sml-github-create-branch |
Create a feature branch following the configured naming convention |
| github-create-pr |
/sml-github-create-pr |
Create a PR from current branch with auto-filled template |
Company Knowledge
Available to all personas.
| Skill |
Command |
What It Does |
Output |
| company-knowledge-connect |
/sml-company-knowledge-connect |
Register internal knowledge sources — framework docs, platform libraries, developer portals — via URL (with optional auth) or a company MCP server |
company_knowledge.sources[] in _superml/config.yml |
| company-knowledge-fetch |
/sml-company-knowledge-fetch |
Pull a registered knowledge source into the current AI session; handles URL auth and MCP tool calls; summarises what was loaded and suggests next actions |
Knowledge in AI context |
Skill Activation Quick Reference
GitHub Copilot
@sml-agent-pm # Activate Product / BA agent
@sml-agent-architect # Activate Architect agent
@sml-agent-developer # Activate Developer agent
@sml-agent-sage # Activate Modernization Lead agent
@sml-agent-lead # Activate Team Lead agent
@sml-agent-scout # Activate Scout agent
/sml-help # What to do next
/sml-create-prd # Write a PRD
/sml-create-architecture # Design architecture
/sml-create-epics-stories # Break into epics and stories
/sml-sprint-planning # Plan a sprint
/sml-dev-story # Implement a story
/sml-code-review # Review code
/sml-relearn-codebase # Onboard to codebase
/sml-read-legacy-code # Deep-read legacy system
/sml-build-knowledge-graph # Extract business rules
/sml-company-knowledge-fetch # Pull internal docs into AI context
Other AI Assistants (Claude, Cursor, etc.)
Load skill at: _superml/skills/2-planning/agent-pm/SKILL.md
Load skill at: _superml/skills/3-solutioning/create-architecture/SKILL.md
Load skill at: _superml/skills/4-implementation/dev-story/SKILL.md