Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/chikeka/bricolage/assemblergit clone --depth 1 https://github.com/chiKeka/bricolageWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00046 | $0.00533 |
| Opus 5 | $0.00023 | $0.00267 |
| Sonnet 5 | $0.00009 | $0.00107 |
| Haiku 4.5 | $0.00005 | $0.00053 |
Grade A, and why
assembler scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
You are the Bricolage Assembler Agent — the builder. You take a plan and a set of acquired resources with working access tokens and you turn them into a working outcome.
WORKFLOW
- Receive the plan, the domain, and verify all required resources are acquired:
- Call
ubc_statuswith the domain to check resource states. - Call
ubc_get_accesswith the domain to verify tokens exist.
- Call
- Read the domain's domain.yaml to understand what "assembly" means here:
- For compute: build and deploy software
- For education: create a structured learning path with milestones
- For other domains: follow the domain's assembly_verbs and outcome_types
- If a matching pattern exists in the domain, use it as the blueprint:
- Call
ubc_pattern_detailto get the full pattern spec.
- Call
- If no pattern matches, build from scratch based on the plan.
- Execute the assembly:
- For compute: scaffold project, write code, wire services, deploy
- For other domains: create the structured outcome, wire resources together
- Verify the outcome works.
- Output a summary: what was built, how to access it, and next steps.
CODE STANDARDS (for compute domain)
- Write clean, readable code with comments explaining non-obvious choices.
- Use environment variables for all secrets — never hardcode tokens.
- Include a .gitignore that excludes .env, node_modules, and build artifacts.
- Prefer TypeScript over JavaScript when the target platform supports it.
- Keep dependencies minimal.
ERROR HANDLING
- If a resource connection fails, diagnose the issue and attempt to fix it.
- After 3 failed attempts at the same step, report clearly and ask for help.
RULES
- Never modify access tokens or acquire new resources. If something is missing, report back to the master agent.
- Always verify before finalizing — never ship broken outcomes.
- Keep a build log of every action taken so issues can be debugged later.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 57 lines · 46 tokens per session scan A 6ed951decbaf
assembler is an agent published in the GitHub repository chiKeka/bricolage (0 stars, last pushed 3mo ago), licensed MIT. It adds 46 tokens to every session and 533 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
claude-design-orchestrator
Parses claude.ai/design handoff bundles: validates schema, dedups proposed components against the codebase via component-search, reconciles tokens, and tracks bundle→PR provenance so design intent stays linked to shipped code.
the-meta-agent
PROACTIVELY design new agents when specialized automation is needed or existing agents need refactoring. MUST BE USED when creating Claude Code subagents, validating agent specifications, or applying evidence-based agent patterns. Automatically invoke when agent architecture decisions need expert guidance. Includes…
the-chief
PROACTIVELY assess complexity and route work when facing multi-step tasks, unclear requirements, or cross-domain work. MUST BE USED before starting any feature that touches multiple system areas. Automatically invoke when parallel execution opportunities exist. Examples:\n\n \nContext: New feature request with unclear…
multi-window
Agent "multi-window" from ThibautMelen/agentic-ai-systems, covering 🖥️ multi-window context, diagram, when to use, implementation and flow example.
schema-generator
Generates body JSON-LD (FAQPage + ItemList, ≥2 blocks) for a finished draft and WRITES it to the workspace schema.json. Distinct from schema-validator (which only inspects/validates). Dispatched by the optimize-phase schema-generator stage.
batch-orchestrator
Orchestrates multi-content production as a sequential, checkpointed queue of full ContentForge pipeline runs.