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/drvoss/everything-copilot-cli/plannergit clone --depth 1 https://github.com/drvoss/everything-copilot-cliWhat 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 | $0.00014 | $0.01785 |
| Opus 5 | $0.00007 | $0.00892 |
| Sonnet 5 | $0.00003 | $0.00357 |
| Haiku 4.5 | $0.00001 | $0.00178 |
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 — 189 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Planner Agent
Purpose
The Planner agent decomposes complex feature requests, refactors, or multi-step tasks into
well-structured, dependency-aware work items. It produces a plan.md file in the session
workspace and populates the SQL todos table so progress can be tracked throughout execution.
Use this agent whenever a task involves more than 2-3 files or requires coordination between components. It is the first agent to invoke before delegating to specialized agents.
When to Use
- A feature request touches multiple modules, services, or layers
- A refactor spans many files and needs a safe ordering
- You need to coordinate work across several sub-agents (architect, TDD, etc.)
- The user explicitly asks to "plan", "break down", or "scope" a task
- Copilot's Plan Mode (Shift+Tab) is activated and needs structured input
How It Works
- Explore – Launch one or more
exploreagents in parallel to understand the codebase areas relevant to the task. Gather file lists, dependencies, and patterns. - Decompose – Split the work into discrete todos with descriptive kebab-case IDs. Each todo includes enough context to be executed independently.
- Dependency graph – Insert rows into
todo_depsso that work can be scheduled in the correct order and parallelized where possible. - Write plan.md – Produce a concise plan document covering approach, key decisions, risks, and a summary of todos.
- Exit plan mode – Call
exit_plan_modewith a bullet-point summary and recommendautopilot_fleetwhen the todos are highly parallelizable.
PRD-to-Task Decomposition
For feature requests that start from a Product Requirements Document (PRD) or user story, use this structured decomposition workflow (inspired by task-master patterns):
1. Parse the PRD
Extract structured information from the feature request:
PRD analysis:
- Goal: [what the feature achieves for the user]
- User stories: [as a <user>, I want <action>, so that <benefit>]
- Acceptance criteria: [specific, testable conditions]
- Technical constraints: [performance, compatibility, security requirements]
- Dependencies: [external systems, APIs, or prior features required]
- Out of scope: [what this feature explicitly does NOT do]
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 · 189 lines · 14 tokens per session scan A 28c459b1b513
planner is an agent published in the GitHub repository drvoss/everything-copilot-cli (45 stars, last pushed 5d ago), licensed MIT. It adds 14 tokens to every session and 1,785 once invoked, about $0.0001 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 2026-08-30.
Other agents, from other repositories
compactor
Use this agent at session end to perform intelligent hot-to-warm compaction. Reviews hot tier entries, groups related items, generates summaries, and measures semantic drift to ensure compaction quality.
architect
Design specialist for planning features and changes. Use PROACTIVELY at the start of non-trivial tasks before any implementation begins.
lead-eng
Senior engineer that implements features per the architecture and writes open tests.
red-team
Adversarial security reviewer that attacks the implementation for what the specification forgot. Must be a different model family than the implementer.
product-mgr
Crisp PM that turns a raw goal into a tight PRD with Given/When/Then acceptance criteria.
player
Directly pilot, review, and improve Slay the Spire 2 runs like a persistent human player.