Skip to content

im-shashanks/claude-plugins / shaktra/skills/shaktra-reference

shaktra-reference

Verified

Shared constants, quality standards, and state schemas for all Shaktra agents. Defines severity taxonomy, quality principles, guard tokens, review dimensions, and YAML schemas.

schemas/briefing-schema.md

View on GitHub
# Briefing Schema

Defines `.shaktra/stories/<story_id>/.briefing.yml` — a per-story filtered view of long-term knowledge, generated at workflow start and consumed by agents during execution.

## Schema

```yaml
story_context:
  id: string             # story ID (e.g., "ST-001")
  keywords: [string]     # extracted from story title, description, and scope

relevant_principles:
  - id: string           # PR-NNN
    text: string         # principle statement
    confidence: float    # current confidence score
    relevance: string    # why this principle matters for this story (1 sentence)
    roles: [string]      # which agents should read this

relevant_anti_patterns:
  - id: string           # AP-NNN
    failed_approach: string  # what went wrong
    recommended_approach: string # what to do instead
    severity: string     # P0-P3
    relevance: string    # why this anti-pattern is relevant
    roles: [string]      # which agents should read this

relevant_procedures:
  - id: string           # PC-NNN
    text: string         # procedure statement
    confidence: float    # current confidence score
    relevance: string    # why this procedure applies
    roles: [string]      # which agents should read this
```

## Generation Rules

### Tier Selection

Determine retrieval tier by calling `memory_retrieval.py` (counts active entries across all memory stores):

| Tier | Condition | Generator |
|---|---|---|
| 1 | total_entries ≤ `settings.memory.retrieval_tier1_max` | Orchestrator generates briefing inline |
| 2 | total_entries ≤ `settings.memory.retrieval_tier2_max` | Spawn single memory-retriever agent (briefing mode) |
| 3 | total_entries > `settings.memory.retrieval_tier2_max` | Spawn parallel chunk retrievers + consolidation retriever |

See `shaktra-memory/retrieval-guide.md` for the full retrieval algorithm and dispatch templates.

### Retrieval Algorithm (all tiers)

1. Read `principles.yml`, `anti-patterns.yml`, `procedures.yml`
2. Extract story keywords from title, description, acceptance criteria, and scope
3. Filter entries by:
   - `status: active` (exclude archived/superseded)
   - `confidence >= settings.memory.briefing_confidence_threshold`
   - Keyword/tag overlap with story context
   - Workflow type match (for procedures)
4. Score relevance: semantic relevance (0-10) × confidence — emphasize semantic matching over keyword overlap
5. Rank descending, cap at `settings.memory.max_briefing_entries`
6. For each matching entry, write a 1-sentence `relevance` explanation
7. Tag each entry with applicable `roles` for agent-filtered reading

## Agent Reading

Agents read the briefing filtered by their role:
- Developer reads entries where `roles` includes "developer"
- SW Quality reads entries where `roles` includes "sw-quality"
- Architect reads entries where `roles` includes "architect"

This ensures agents receive only relevant knowledge without loading the full memory stores.
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall