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 skills/johnny-xuan/treework/treework-manualnpx skills add Johnny-xuan/TreeWork --skill treework-manualgit clone --depth 1 https://github.com/Johnny-xuan/TreeWorkWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/johnny-xuan/treework/treework-manual)<a href="https://agentmods.dev/skills/johnny-xuan/treework/treework-manual"><img src="https://agentmods.dev/badge/skills/johnny-xuan/treework/treework-manual.svg" alt="Measured on agentmods" height="20"></a>What 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 5.1 | $0.00068 | $0.02355 |
| Opus 5 | $0.00034 | $0.01177 |
| Sonnet 5 | $0.00014 | $0.00471 |
| Haiku 4.5 | $0.00007 | $0.00235 |
Grade A, and why
treework-manual 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 6d 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 — 293 lines — stays where its author put it; the contents beside it link to each section on GitHub.
TreeWork Manual
TreeWork gives an Agent a durable sense of place in ongoing work. It externalizes the work as a Tree: the root holds the whole project, each branch owns one coherent line of effort, and documents at the root and branches preserve intent, reality, and durable knowledge. Direction no longer depends on the current conversation remaining intact.
The Agent's unit of movement is not the next message; it is the current branch. A new request is first located in the Tree. The Agent then works from the root into the branch that owns the work, leaves that branch recoverable, returns to the root, and chooses the next branch. Requests may arrive in any order, but the work still moves through the Tree.
Start at the root
-> locate the right branch
-> work within that branch
-> leave a recoverable state
-> return to the root
-> choose the next branch
The Tree provides direction; the documents provide continuity:
PROJECT shows the global map.
Plan points toward intended work.
Progress records present reality.
Findings preserve what the work has taught the Agent.
TreeWork is a working mental model rather than a prescribed sequence. A branch may be a chapter, theme, research question, note collection, deliverable, phase, module, client, or another coherent scope that fits the project. The Tree may begin shallow and grow as understanding improves. The Agent maintains this topology and its state directly in Markdown so that people and other Agents can inspect, correct, and continue the same work.
Core Mental Model
- Locate before acting. Route each meaningful request to the root, the current branch, another existing branch, or a genuinely new branch before doing the work.
- Keep global and local state separate. Keep the root concise and navigational. Keep details inside the branch that owns them.
- Move instead of jumping. When attention must move elsewhere, first leave the current branch in a state that can be resumed without conversation history.
- Let reality correct the documents. Plans express intent; Progress records what is actually true. When artifacts, user direction, and older documents disagree, investigate and synchronize them instead of preserving a stale story.
- Grow the Tree only as needed. Sketch enough structure to orient the work, then refine it as understanding improves. Do not invent a large branch hierarchy merely to appear organized.
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.
- 6d ago First seen · 293 lines · 68 tokens per session scan A 66d0ffc5c2cb
treework-manual is a skill published in the GitHub repository Johnny-xuan/TreeWork (14 stars, last pushed 24d ago), licensed MIT. It adds 68 tokens to every session and 2,355 once invoked, about $0.0003 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 skills, from other repositories
repository-bootstrap
Audit and establish or upgrade the canonical repository governance baseline. Use only when explicitly asked to bootstrap, standardize, or upgrade a repository; do not use for feature development or source-code restructuring.
repository-hygiene
Perform a read-only evidence-based audit of stale docs, agent files, scripts, generated artifacts, duplicate code, and optional tools. Use only when explicitly asked for cleanup or governance review; do not delete or refactor during the audit.
safe-change
Trace, implement, test, and independently review one bounded software change. Use for ordinary feature, bug-fix, or refactor work after the repository is bootstrapped; do not use for broad repository cleanup, production operations, or undefined rewrites.
owner-driven-development
Execute an approved plan with one persistent implementation owner, explicit Git disposition, bounded delegation, and risk-triggered review.
risk-adaptive-review
Classify risk deterministically, review complete working-tree changes at architectural seams, and converge with one recheck.
verification-before-completion
Discover canonical project gates and require fresh, correctly classified evidence before any completion claim.