Assemble relevant context packages for agent spawns within token budgets.
npx skills add https://github.com/mehdic/CDC --skill context-assembler
Please help me install this Agent Skill. Skill: context-assembler Source: https://github.com/mehdic/CDC/tree/main/.claude/skills/context-assembler Command: npx skills add https://github.com/mehdic/CDC --skill context-assembler
This Skill compiles the most relevant context packages for a given agent spawn, prioritizing by relevance and staying within token budgets.
Assemble context for a Developer spawn in session bazinga_20250112_143022 and write results to bazinga/artifacts/{SESSION_ID}/skills/context.json.
High-intent search queries and answers about installing and using this skill.
Context assembler retrieves and prioritizes relevant context packages for agent spawns while enforcing token budget limits. It ranks packages by relevance, applies per-agent retrieval limits, and outputs structured context blocks ready to embed in downstream prompts without exceeding your token allocation.
The Skill applies graduated token budget zones and per-agent limits loaded from configuration for developer, senior engineer, QA, tech lead, and investigator roles. Each role receives prioritized context packages up to its configured retrieval limit, ensuring balanced token distribution across agent spawns.
Yes, context assembler uses FTS5-agnostic heuristic ranking to prioritize packages by relevance without requiring full-text search infrastructure. It applies configurable ranking logic and token budget calculations with safety margins, then outputs ranked context blocks for immediate use.
The Skill retrieves error patterns alongside context packages and reasoning data during orchestrator-driven agent spawns. It captures error patterns for learning, enabling downstream agents to refine their reasoning based on historical failures and improving future context selection.
Provide context packages, error patterns, reasoning data, and per-agent configuration specifying retrieval limits and token budgets. The Skill processes these inputs during an orchestrator-driven agent spawn and outputs structured context blocks to a configurable artifact path.
Heuristic ranking eliminates dependency on full-text search infrastructure while maintaining relevance prioritization and token budget enforcement. It scales efficiently for agent-driven workflows, applies configurable safety margins to token calculations, and outputs context blocks immediately without external indexing overhead.
TravisLeeeeeeConvert raw email threads into structured JSON context with citations.
89jobrienOptimize LLM prompts with structured design, few-shot patterns, and chain-of-thought reasoning.
mehdicExecute prompt_builder.py to generate agent prompts with token budgets.
JhabbigImplement similarity search patterns for vector databases and ranked retrieval.
