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 skills add pome-sh/digital-twins --skill pome-suggest-tasksgit clone --depth 1 https://github.com/pome-sh/digital-twinsWrote 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/pome-sh/digital-twins/pome-suggest-tasks)<a href="https://agentmods.dev/skills/pome-sh/digital-twins/pome-suggest-tasks"><img src="https://agentmods.dev/badge/skills/pome-sh/digital-twins/pome-suggest-tasks/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/pome-sh/digital-twins/pome-suggest-tasks"><img src="https://agentmods.dev/badge/skills/pome-sh/digital-twins/pome-suggest-tasks.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00041 | $0.00923 |
| Opus 5 | $0.00020 | $0.00462 |
| Sonnet 5 | $0.00008 | $0.00185 |
| Haiku 4.5 | $0.00004 | $0.00092 |
Grade A, and why
pome-suggest-tasks 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 8d 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 — 74 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Pome suggest tasks (own-agent front door)
You are the coach: you talk to the builder and to the Pome control MCP
(mcp.pome.sh). The examinee is the builder's own agent, already registered
(one pome.json, pome.yaml, or pome.yml is present). This skill turns a registered-but-untested agent into
one chosen candidate task by reading what the agent actually does and
proposing grounded candidates. It authors nothing: drafting, validating, and
running belong to pome-author-task → pome-verify-seed → pome-run-task.
Runs after registration, before Skill 1. If there is no Pome manifest, the agent is not registered yet — route back to the entry path, do not improvise a registration here.
If the mcp__pome__* tools are missing, the MCP isn't connected: ask the user to
connect and authenticate it (interactive OAuth — needs a human in a browser)
instead of probing the endpoint. You can still read the agent and propose
candidates offline; the catalog check in step 2 just waits for the connection.
1. Read the agent
Read the manifest and the code — this is the ground truth every candidate must stand on:
- The Pome manifest —
agent.slug,twins(the ONLY surfaces a candidate may exercise),command, and thetasksdir. - The agent's prompt + entry code — the system prompt and the tool set (e.g.
src/index.ts). What actions can it actually take, and what does the prompt tell it to do — or, tellingly, NOT do? An unstated check (no authorization step, no severity rule) is exactly where a fear lives.
Name, per twin, the concrete actions the agent can take. A candidate that touches
a twin not in twins is out of scope — drop it before you propose it.
2. Propose 2–3 candidates grounded in the agent
Author from fear, not from features (the pome-author-task framing): each
candidate is one bad end-state the agent could reach and the good one it should.
Ground every candidate in something you actually read — a missing check in the
prompt, a tool it holds, a surface it trusts — never a generic template.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 8d ago Changed · -71 tokens per session 85cd53946ace
- 12d ago First seen · 74 lines · 112 tokens per session scan A ef0fb3d67355
pome-suggest-tasks is a skill published in the GitHub repository pome-sh/digital-twins (14 stars, last pushed 6d ago), licensed Apache-2.0. It adds 41 tokens to every session and 923 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 2026-08-30.
Other skills, from other repositories
cw-gates
Use before claiming any Codewhale change is done, green, or ready to land: the focused-to-broad verification ladder, the budget checks CI enforces, and the rules for what counts as a passing test.
feishu
Work with Feishu or Lark bots, docs, sheets, bitables, approval flows, and OpenAPI/MCP setup without hardcoding credentials.
interview
Ask one useful structured question at a time only when material product/implementation choices are genuinely missing; remember answers and produce a brief/spec. Discoverable facts should be investigated instead of asked.
ppt-review
Read this skill only after current officereview structural output and relevant rendered images exist. Do not use source code, a tool success flag, or the first-slide preview as a substitute for deck evidence.
package-installer
Install third-party open-source projects (a GitHub URL or a local git directory the user provides) as external packages. The project is cloned verbatim into the per-user packages directory — never edit, normalize, or reorganize its files.
cog-knowledge-consolidation
Build structured knowledge frameworks from scattered vault notes with source attribution.