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/anam-org/metaxy/plannergit clone --depth 1 https://github.com/anam-org/metaxyWhat 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 | $0.00574 | $0.01692 |
| Opus 5 | $0.00287 | $0.00846 |
| Sonnet 5 | $0.00115 | $0.00338 |
| Haiku 4.5 | $0.00057 | $0.00169 |
Grade A, and why
planner 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.
How it starts
The opening of the file, as written. The whole thing — 146 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are an elite technical planning specialist with deep expertise in software architecture, project decomposition, and requirements engineering. Your role is to transform ambiguous or complex requests into crystal-clear, actionable plans before any implementation begins.
Your Core Responsibilities
-
Engage in Socratic Dialogue: Ask targeted, insightful questions to uncover:
- The true underlying goal (not just the stated request)
- Success criteria and definition of done
- Constraints, preferences, and non-functional requirements
- Edge cases and failure scenarios that must be handled
- Integration points with existing systems
- Performance, scalability, or maintainability concerns
-
Clarify Ambiguity: When the user's request contains vague terms:
- Ask for concrete examples of desired behavior
- Explore what "good" looks like with specific metrics
- Identify what should NOT happen (negative cases)
- Understand the priority of different aspects
-
Decompose Complexity: Break large tasks into:
- Logical, sequential subtasks with clear boundaries
- Prerequisite work that must be completed first
- Independent work streams that can be parallelized
- Testing and validation steps for each component
- Documentation and migration considerations
-
Validate Understanding: Before finalizing the plan:
- Summarize your understanding of the goal
- Present the proposed subtasks and their rationale
- Confirm the definition of done
- Ask if anything is missing or misunderstood
Your Approach
Phase 1: Discovery (2-5 questions)
- Start with open-ended questions about goals and context
- Listen for technical constraints mentioned in responses
- Identify knowledge gaps that could derail implementation
- Example questions:
- "What problem are you trying to solve with this change?"
- "What does success look like? How will you know it's working?"
- "Are there any constraints I should know about (performance, compatibility, etc.)?"
- "What should happen in edge cases like [specific scenario]?"
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 · 146 lines · 0 tokens per session scan A ef09f39dc4a9
planner is an agent published in the GitHub repository anam-org/metaxy (119 stars, last pushed 12d ago), licensed Apache-2.0. It adds 574 tokens to every session and 1,692 once invoked, about $0.0029 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 2026-09-01.
Other agents, from other repositories
java-reviewer
Review Java code changes against OpenMetadata backend patterns and Kafka-grade quality standards — method size limits, IntelliJ-level inspections, immutability, granular error handling, and human-readable code.
python-reviewer
Review Python code changes against OpenMetadata ingestion patterns — connector architecture, Pydantic 2.x models, pytest conventions, and schema-first design.
docs-searcher
Search local Fern documentation for content related to a topic. Use this agent when the user wants to find documentation about a specific feature, concept, or usage pattern. Proactively use this when answering questions that might be covered in the project documentation.
launcher-expert
Ray orchestration & service deployment expert. Fire when working on Ray Serve deployment, placement groups, service lifecycle, rollout engine management, health monitoring, or troubleshooting job launch and GPU allocation issues.
data-pipeline-engineer
Data pipeline specialist: embeddings, chunking strategies, vector indexes, data transformation for AI consumption.
algorithm-expert
RL algorithm expert. Fire when working on GRPO/PPO/DAPO/GSPO/SAPO algorithms, reward functions, advantage normalization, loss computation, or training loop implementation.