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 commands/pcvelz/cloudone/convert-existing-commandgit clone --depth 1 https://github.com/pcvelz/cloudoneWhat 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 | $0.00058 | $0.01524 |
| Opus 5 | $0.00029 | $0.00762 |
| Sonnet 5 | $0.00012 | $0.00305 |
| Haiku 4.5 | $0.00006 | $0.00152 |
Grade A, and why
convert-existing-command 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 — 145 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/convert-existing-command
$ARGUMENTS names the slash command to convert, and the name for the new
sibling command: <source> <new-name>.
The output is a second command. The source file is never touched — not rewritten, not backed up, not moved. If both arguments are not present, ask; never guess a name and never write over an existing file.
Read .claude/docs/delegation-contract.md before you start. This command
produces artifacts defined there (the state machine, the result schema, the
contract block) and must not restate or reinvent them.
1. Locate the source
- If the first argument resolves to a file, use it.
- Otherwise search
./.claude/commands/*.mdand~/.claude/commands/*.mdfor a matching name. Multiple or zero matches → list what you found and ask which file to convert. - If the second argument is missing, ask for the sibling name. If a file with that name already exists, stop and ask — never overwrite.
2. Suitability check
Read the source command fully. Delegation fits when the body contains self-contained batch work: running a pipeline or test suite, analysing a diff, generating files, bulk edits.
It fits badly when the body is dominated by things that need the calling session: reading conversation context, asking the user questions mid-flow, or producing results the caller must act on immediately.
Badly-suited is not the same as impossible — that is what the split in the next section is for. Say in two sentences which parts look coupled to the caller, and let the split table resolve it. Only abandon the conversion if nothing substantial is left to delegate once the coupled steps are removed.
3. Architectural phase (before writing anything)
The communication design is fixed here, once, and then embedded in the emitted command. It is never decided at runtime. Produce three artifacts and show all three to the user for sign-off before emitting.
(a) The split table. Walk every step of the source body and answer one question: does this step need the caller's identity or conversation state? Yes → retained. No → delegated. Identity-bearing work includes anything stamped with the caller's session id, anything that reads the conversation, and anything the caller must act on immediately.
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 · 145 lines · 58 tokens per session scan A 02a7d9a4a3d3
convert-existing-command is a command published in the GitHub repository pcvelz/cloudone (1 stars, last pushed 19d ago), licensed MIT. It adds 58 tokens to every session and 1,524 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 2026-08-31.
Other commands, from other repositories
netllm-connect
Wire Cursor, Claude Code, Codex, or Honcho to the local netllm router.
netllm-swarm
Configure multi-machine LAN mesh for netllm (mDNS, peers, gateway).
netllm-doctor
Troubleshoot netllm misconfigurations (PATH, providers, mDNS, agent health).
netllm-setup
First-time netllm install from this repo (uv sync, init, discover, verify).
forget
Delete a memory permanently. Asks for confirmation first.
graph
Show memories related to a given one — knowledge-graph traversal up to 3 hops.