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 Hmbown/Wizards-of-the-Ghosts --skill eyebitegit clone --depth 1 https://github.com/Hmbown/Wizards-of-the-GhostsWrote 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/hmbown/wizards-of-the-ghosts/eyebite)<a href="https://agentmods.dev/skills/hmbown/wizards-of-the-ghosts/eyebite"><img src="https://agentmods.dev/badge/skills/hmbown/wizards-of-the-ghosts/eyebite/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/hmbown/wizards-of-the-ghosts/eyebite"><img src="https://agentmods.dev/badge/skills/hmbown/wizards-of-the-ghosts/eyebite.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 | $0.00106 | $0.00770 |
| Opus 5 | $0.00053 | $0.00385 |
| Sonnet 5 | $0.00021 | $0.00154 |
| Haiku 4.5 | $0.00011 | $0.00077 |
Grade A, and why
eyebite 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Eyebite
Fix your attention on a target and selectively degrade its capabilities.
What This Skill Does
In D&D, Eyebite lets you focus on one creature per turn and inflict sleep, panic, or sickness through sustained eye contact. The real-world version is targeted capability reduction: focused analysis that identifies and disables specific functions of a system, service, or adversary. Feature flagging a dangerous capability off. Selectively throttling a misbehaving API consumer. Disabling specific attack vectors during an incident. Eyebite requires sustained focus on a single target — it is not a broadcast weapon. In this grimoire, Eyebite is treated as a metaphorical spell with a prototype delivery profile. Canonical reference input: Eyebite (spell).
When To Use
- A specific system, service, or feature needs to be selectively degraded rather than fully shut down.
- You are responding to an incident and need to disable specific capabilities while keeping the rest operational.
- Targeted throttling or feature disabling is more appropriate than a full kill or freeze.
Prerequisites
- No extra runtime dependencies beyond Hermes Agent and the normal toolset for this session.
Procedure
- Restate the target, the success condition, and any no-touch boundaries before taking action.
- Identify the specific target and the capability you want to degrade or disable.
- Assess dependencies — what else relies on the capability you are about to remove?
- Execute the selective degradation: feature flag, targeted throttle, capability restriction.
- Monitor the target to confirm the degradation is working as intended without cascading side effects.
- Document the intervention and set a review point for restoration.
- Package the result as the deliverables below, with confidence, assumptions, and unresolved risk called out explicitly.
Deliverables
- Confirmation of which capabilities were degraded and how.
- A dependency impact assessment.
- A restoration plan for when the degradation should be reversed.
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 · 70 lines · 106 tokens per session scan A 37f5e73aa3b8
eyebite is a skill published in the GitHub repository Hmbown/Wizards-of-the-Ghosts (107 stars, last pushed 5mo ago), licensed CC0-1.0. It adds 106 tokens to every session and 770 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 2026-09-03.
Other skills, from other repositories
linter-agent
Use when detecting and fixing code style violations, enforcing project conventions, ensuring consistent formatting.
perf-agent
Use when measure before optimizing, target actual bottlenecks proven by profiling, verify with benchmarks.
manage-skills
A maintenance workflow for checking whether project verification skills still cover the code and rules that changed during a session.
systematic-debugging
Structured debugging methodology — use before proposing fixes for any error or failure. Covers: code bugs, build errors, deploy failures, config conflicts, dependency issues, infra problems. Also use when previous fix attempts failed or root cause is unclear.
review-loop
Run the adversarial verification loop — implement, then hand the change to a fresh checker that did not write it, fix what it finds, and re-dispatch until APPROVE. Use before claiming any behavioural change is done, and on requests like "review loop", "adversarial review", "independent review", "get this verified"…
reproduce
Reproduce a cloudflare/agents GitHub issue by scaffolding a minimal Agents/Worker project and deploying it to a temporary Cloudflare account, then report findings back on the issue.