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 skills/ikarusmk/aicortex/secret-handlingnpx skills add IkarusMK/AIcortex --skill secret-handlinggit clone --depth 1 https://github.com/IkarusMK/AIcortexWrote 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/ikarusmk/aicortex/secret-handling)<a href="https://agentmods.dev/skills/ikarusmk/aicortex/secret-handling"><img src="https://agentmods.dev/badge/skills/ikarusmk/aicortex/secret-handling.svg" alt="Measured on agentmods" 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 | $0.00034 | $0.00319 |
| Opus 5 | $0.00017 | $0.00160 |
| Sonnet 5 | $0.00007 | $0.00064 |
| Haiku 4.5 | $0.00003 | $0.00032 |
Grade A, and why
secret-handling 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 5d 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.
What it actually says
Secret Handling
When to use
Any time an API key, token, or password appears or is needed.
The rule (hard)
Secrets go in the encrypted vault, immediately, referenced by name — never pasted into chat, committed to a repo, written into a memory, or hardcoded.
How
- Store:
secret_set("SOME_API_KEY", <value>)→ encrypted at rest, never returned. - Reference: a service/device points at the name (
token_env="SOME_API_KEY"), and the server injects the value at call time. The model never sees it. - List/remove:
secret_list(names only) ·secret_delete(name).
Don't
- Don't ask the user to paste a secret "just to check it".
- Don't put a secret in a service config, a memory, or a commit — those are not encrypted.
- Don't echo a secret back, even partially.
If a secret leaks (chat, repo, logs)
Treat it as compromised: rotate it at the source, update the vault with the new value, and remove the exposed copy. A leaked encryption key means re-encrypting the vault — rotate carefully and re-store the affected secrets.
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.
- 5d ago First seen · 32 lines · 34 tokens per session scan A 43728c8ed97d
secret-handling is a skill published in the GitHub repository IkarusMK/AIcortex (9 stars, last pushed 16d ago), licensed Apache-2.0. It adds 34 tokens to every session and 319 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 2026-08-31.
Other skills, from other repositories
unraid
This skill should be used when the user mentions Unraid, asks to check server health, monitor array or disk status, list or restart Docker containers, start or stop VMs, read system logs, check parity status, view notifications, manage API keys, configure rclone remotes, check UPS or power status, get live CPU or…
swag
This skill should be used when the user says "add proxy config", "create reverse proxy", "SWAG config", "nginx proxy", "expose service", "proxy configuration", "subdomain config", "subfolder config", "configure SWAG", "list proxy configs", "view proxy config", "edit proxy config", "remove proxy config", "check proxy…
using-human-dom
Use when you need DOM-level precision (exact text/aria/placeholder matching → screen coordinates) while operating the host's real Chrome with human identity and zero automation traces. humandom adds a read-only DOM locator alongside humanbrowser; all actions remain OS-level (tap/typetext/presskey). When the DOM…
using-mac
Use when invoking mac-device MCP tools to drive a Mac test machine (agent-fleet project) -- screenshot-and-click UI testing, long-running shell or zsh commands, file or process operations, AppleScript automation, or coordinating exclusive device use across multiple agents.
using-win
Use when invoking win-device MCP tools to drive a Windows test machine (agent-fleet project) -- screenshot-and-click UI testing, long-running shell commands, file or process operations, or coordinating exclusive device use across multiple agents.
using-vision
Use when an element you need to click is NOT in the OS accessibility tree — web pages, canvas, Electron/Flutter apps, games — so findelements/tapelement return nothing. Provides pixel-level location by visible text (OCR) or by icon image (template match), returning coordinates in the same space as tap. mac/win…