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 trentisiete/endy --skill endy-delegategit clone --depth 1 https://github.com/trentisiete/endyWrote 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/trentisiete/endy/endy-delegate)<a href="https://agentmods.dev/skills/trentisiete/endy/endy-delegate"><img src="https://agentmods.dev/badge/skills/trentisiete/endy/endy-delegate/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/trentisiete/endy/endy-delegate"><img src="https://agentmods.dev/badge/skills/trentisiete/endy/endy-delegate.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 | $0.00104 | $0.04981 |
| Opus 5 | $0.00052 | $0.02491 |
| Sonnet 5 | $0.00021 | $0.00996 |
| Haiku 4.5 | $0.00010 | $0.00498 |
Grade A, and why
endy-delegate 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 9d 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 — 348 lines — stays where its author put it; the contents beside it link to each section on GitHub.
endy-delegate
Hand off a coding task to an endy subagent. Three delivery modes:
- short — blocking bash call, you wait for the output.
- long — detached tmux window, you get a task ID back and check on it later.
- handoff — transfer an in-flight task from one agent to another (e.g. opencode ran out of rate limit; cmd picks up with the same prompt + opencode's full log).
Handoff: continue an in-flight task with a different agent
endy handoff <task-id> is the cross-agent transfer. It exists for the
case "this agent can't finish the task — get a different one to pick up
without losing context".
endy handoff <task-id> --to <next-agent> --reason "<short>" [--stop-parent]
What it does, automatically:
- Reads the parent task's
.meta, full.log, andprompt.md. - Composes a new prompt with explicit handoff markers:
[endy handoff — you are taking over from a previous agent]- The original task prompt verbatim.
- The previous agent's FULL output (clear-to-EOL, ANSI-stripped). Cap
it with
--lines Nonly when the target has a small context window (e.g. gemini free). - The reason you provided.
- Spawns the new agent in the same tmux session as the parent. Inherits cwd and orchestrator label.
- Records the chain:
handoff_from,handoff_chain(multi-hop accumulates),handoff_reasonin the new task's meta. - With
--stop-parent: closes the rate-limited window in the same shot.
If ENDY_HANDOFF_RESOLVER=multiplexor-next-provider is set (it is, by
default after endy install), drop --to entirely — multiplexor picks
the next eligible agent automatically:
endy handoff <task-id> --reason "rate limited"
# → multiplexor marks the previous agent exhausted, returns next-best
# eligible (e.g. opencode → cmd), endy spawns it with the composed
# prompt above.
When to call handoff (vs. spawn / live)
| Situation | Use |
|---|---|
| Fresh task, no in-flight work to inherit | endy spawn <agent> |
| Fresh task you want to drive interactively | endy live open (see endy-live skill) |
| In-flight task hit a rate limit / quota / auth error | endy handoff <task-id> |
| Same agent, just continue the conversation | endy watch followup <id> (native resume) |
| Different agent, take over what was being done | endy handoff <id> --to <agent> |
What ships with it
1 file 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.
- 9d ago First seen · 348 lines · 104 tokens per session scan A 10d038b657dd
endy-delegate is a skill published in the GitHub repository trentisiete/endy (7 stars, last pushed 3mo ago), licensed MIT. It adds 104 tokens to every session and 4,981 once invoked, about $0.0005 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 skills, from other repositories
go
Execute an approved Hames task contract when the user invokes /go.
setup
Configure or diagnose Hames in the current project when the user invokes /setup.
ready
Prepare and approve a bounded task contract when the user invokes /ready.
doctor
Diagnose a Hames installation without changing it when the user invokes /doctor.
context-rot-cleanup
Examine workspace files for durable context, filter out low-value noise, and clean up stale memory or rule files that contribute to context rot.
commander-memory-cleanup
Distill Herd commander memory, clean stale MEMORY.md entries, promote durable commander lessons into LONGTERMMEM.md, and propose shared-knowledge updates through a review gate. Use for periodic commander memory cleanup.