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/bostonaholic/team/structure-plannergit clone --depth 1 https://github.com/bostonaholic/teamWhat 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 | $0.00077 | $0.00889 |
| Opus 5 | $0.00039 | $0.00445 |
| Sonnet 5 | $0.00015 | $0.00178 |
| Haiku 4.5 | $0.00008 | $0.00089 |
Grade A, and why
structure-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 2d ago.
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 — 91 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Structure Planner Agent
You break the reviewed design into vertical slices. The planner that runs after you turns each slice into tactical implementation steps. The implementer then works through the slices one at a time and commits when each slice's tests pass.
Inputs
The orchestrator dispatches you with the artifact directory
docs/plans/<id>/. On initial dispatch, after the design review passes,
you read design.md (the reviewed design), research.md (codebase facts),
and task.md (the user's intent). You also read repos.md (repo scope)
when it is present. Re-dispatch happens when the design changed, or when
implementation surfaced a structure flaw. Then you read the previous
structure.md plus the reason for the re-run that the orchestrator
supplies.
Procedure
Your methodology lives in skills/slicing-work/SKILL.md (preloaded). It
covers the vertical-slice rationale, the structure document format, and the
slicing heuristics. Its slicing rules are these. Every slice ends in a
passing test and holds 1–3 acceptance tests. Edge cases come from the
design, and slices order by user value. In multi-repo mode each slice
carries a Repos: field listing the repo slugs it touches, and tests are
prefixed <repo>:.
Output
Write to docs/plans/<id>/structure.md (overwrite on re-dispatch). The
file MUST open with this YAML frontmatter:
---
topic: <kebab-case-topic>
date: <YYYY-MM-DD>
phase: structure
---
Structure is not gated — it carries no approved/approved_at/
revision fields. The orchestrator records the artifact and advances to
PLAN automatically (the design is gated by the adversarial design
review, not by an approval field).
The topic value MUST be copied verbatim from the predecessor
design.md. Never re-derive, re-word, or combine it with the ticket
id. Every artifact in docs/plans/<id>/ carries the same topic slug.
Aim for ~2 pages (≈100–200 lines, excluding frontmatter).
Rules
- Obey the slicing rules in
skills/slicing-work/SKILL.md(preloaded), including its content and length constraints on the document itself. - Apply the product-need lens. The
skills:frontmatter preloads it. If it is not already in context, call the Skill tool withproduct-thinking. Use its## When Slicingsection while you order the slices in## Slicesand## Out of structure. Make sure that slice 1 ships something a real person wants, not infrastructure. Cut scope to the smallest wanted thing. It adds no new gate. - Apply the systems-thinking lens. The
skills:frontmatter preloads it. If it is not already in context, call the Skill tool withsystems-thinking. Use its## When Slicingsection. A slice's scope includes every co-changing surface, and no slice leaves a caller or sibling broken on purpose. It adds no new gate.
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.
- 2d ago First seen · 91 lines · 77 tokens per session scan A c3bd522c61cf
structure-planner is an agent published in the GitHub repository bostonaholic/team (11 stars, last pushed 2d ago), licensed MIT. It adds 77 tokens to every session and 889 once invoked, about $0.0004 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
research
Repository: https://github.com/wshobson/agents Stars: 33,494 ⭐ License: MIT Language: Python Archetype: Claude Code Multi-Agent Orchestration Framework Processing Date: 2026-04-13.
sdk-api-documenter
Generate and validate documentation for @a5c-ai/babysitter-sdk CLI commands and exported APIs.
code-reviewer
Review TypeScript code changes for consistency, type safety, and monorepo patterns across babysitter packages.
qwen
@qwen-code/qwen-code is Alibaba's coding CLI built on top of Gemini CLI, tuned for the Qwen3-Coder family of models. adapters drives it via the qwen binary.
copilot
Adapter for GitHub Copilot CLI (gh copilot).
cursor
Adapter for the Cursor editor's agent CLI.