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 allemaar/open-skills --skill agent-outputgit clone --depth 1 https://github.com/allemaar/open-skillsWrote 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/allemaar/open-skills/agent-output)<a href="https://agentmods.dev/skills/allemaar/open-skills/agent-output"><img src="https://agentmods.dev/badge/skills/allemaar/open-skills/agent-output/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/allemaar/open-skills/agent-output"><img src="https://agentmods.dev/badge/skills/allemaar/open-skills/agent-output.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 | $0.00054 | $0.00605 |
| Opus 5 | $0.00027 | $0.00302 |
| Sonnet 5 | $0.00011 | $0.00121 |
| Haiku 4.5 | $0.00005 | $0.00060 |
Grade A, and why
agent-output 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 12d 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/agent-output
An agent message exists to change the receiver's state. Send only information needed to decide, verify, continue, or stop.
Default report envelope
Omit empty fields. Keep each field to one compact block.
STATE: done | partial | blocked
RESULT: <verdict or completed outcome>
EVIDENCE: <exclusive proof, exact source, or check result>
CHANGED: <paths, symbols, artifacts, or none>
GAPS: <not checked, uncertainty, or none>
NEXT: <one required action or none>
For a finding, use FINDING, IMPACT, EVIDENCE, FIX, and GAPS. For a
short live update, use only STATE plus the changed fact. Do not send the full
envelope when two lines carry the message.
Compression rules
- Lead with the result. Do not restate the assignment.
- Report effects and evidence, not work chronology.
- Name exact artifacts. Do not paste large diffs, logs, or source bodies.
- Combine repeated findings by root cause. Preserve distinct severities.
- State uncertainty and untested scope once, where they affect reliance.
- Use stable identifiers as trailing anchors. Never make the receiver decode identifiers to learn the point.
- Send one recommendation or next action. More than one only when the decisions are independent.
- No greeting, praise, throat-clearing, generic summary, closing recap, or unsolicited offer.
- Do not repeat a fact across
RESULT,EVIDENCE, andNEXT.
Fidelity floor
Brevity never removes authority limits, stop conditions, destructive-action gates, exact numbers, scope, uncertainty, attribution, failed checks, or what was not checked. A short report that conceals a reversing caveat is invalid.
Dispatch use
When assigning work, include this compact return contract in the worker brief. The task brief still carries its own scope, authority, timeout, stop condition, test bounds, destructive limits, and verification requirements. This skill compresses the report. It does not weaken the assignment.
Boundary
- Not
/agent-mailbox. That skill transports durable messages. - Not
/handoff. That skill builds a cold-agent context package. - Not
/human-output. That governs Handler-facing prose. This contract is for agents who need dense operational signal.
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.
- 12d ago First seen · 70 lines · 54 tokens per session scan A df8fcb2ccbe2
agent-output is a skill published in the GitHub repository allemaar/open-skills (14 stars, last pushed 2d ago), licensed Apache-2.0. It adds 54 tokens to every session and 605 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
polis-protocol
Set up and run a self-improving, multi-vendor AI agent team with the Polis Protocol — a markdown polis/ folder where each agent is a citizen with a capability card, tasks are contracts routed to whoever has the best track record by a learning router, settled work files lessons that compound into team memory, and…
media
Skill "media" from ymxlx/polis-protocol, covering polis protocol: a self-optimizing city of agents, the core idea, when this skill is active, structure of a polis and the first thing to do every session.
_data
Skill "_data" from ymxlx/polis-protocol, covering polis protocol: a self-optimizing city of agents, the core idea, when this skill is active, structure of a polis and the first thing to do every session.
subagent-driven-development
Execute plans via delegatetask subagents (2-stage review).
mcporter
List, auth, and call MCP servers/tools from the terminal.
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.