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 impactbrussels/AINativeOS --skill secure-the-connectorsgit clone --depth 1 https://github.com/impactbrussels/AINativeOSWrote 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/impactbrussels/ainativeos/secure-the-connectors)<a href="https://agentmods.dev/skills/impactbrussels/ainativeos/secure-the-connectors"><img src="https://agentmods.dev/badge/skills/impactbrussels/ainativeos/secure-the-connectors/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/impactbrussels/ainativeos/secure-the-connectors"><img src="https://agentmods.dev/badge/skills/impactbrussels/ainativeos/secure-the-connectors.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 | $0.00185 | $0.01534 |
| Opus 5 | $0.00093 | $0.00767 |
| Sonnet 5 | $0.00037 | $0.00307 |
| Haiku 4.5 | $0.00018 | $0.00153 |
Grade A, and why
secure-the-connectors 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 8d 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 — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Secure the Connectors
The model is harmless until you hand it a tool. Every connector you wire, an MCP server, an API key, a file store, a payments endpoint, widens what an injected instruction can reach, and the agent cannot tell your instruction from a hostile one buried in a document it just read. Most agent incidents are not the model "going rogue"; they are a poisoned input firing a tool that was granted more access than the job ever needed. Securing the connectors is deciding, before access is granted, what each tool may touch and what a human must sign for, so an injection that wins still cannot do harm.
The method
Inventory the tools, scope each to the minimum, and gate the dangerous ones behind a human. Full framework: references/connector-hardening.md. Sibling: red-team-the-agent attacks the running agent; this bounds the tools so the attack has nothing to grab.
Step 1: Inventory every connector and its blast radius
List each tool the agent can call and what it can actually reach. You cannot secure a surface you have not drawn.
| Connector | What it can reach | Reversible? | Blast radius if abused |
|---|---|---|---|
| File store (read) | Private docs in scope X | n/a (read) | Data exfiltration via a crafted prompt |
| Email / send | Anyone | No | Sends on the agent's authority |
| Payments / refund | Customer balances | No | Unbounded payout |
| Database (write) | Production rows | No | Silent corruption |
Step 2: Classify read-only versus irreversible, and enforce scope server-side
Sort every tool into read-only or state-changing, and irreversible or recoverable. Then grant the minimum, and grant it where the agent cannot talk its way out.
| Tool class | Default grant | Where it is enforced |
|---|---|---|
| Read-only, non-sensitive | Allow | Connector scope |
| Read-only, sensitive data | Allow only the needed records | Server-side filter, not the prompt |
| State-changing, reversible | Allow with an audit log | Connector + audit trail |
| Irreversible (pay, delete, send) | Deny by default | Human approval, server-side |
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.
- 8d ago First seen · 90 lines · 0 tokens per session scan A 7fd98e70cad7
secure-the-connectors is a skill published in the GitHub repository impactbrussels/AINativeOS (1 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 185 tokens to every session and 1,534 once invoked, about $0.0009 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
remember
Review reusable project knowledge and decide what belongs in project memory, notepad, or durable docs.
docx-comment-reply
Reply to comments (批注) in Word .docx/.doc files: extract comment context, draft replies, write threaded replies back, and validate OOXML.
explaining-machine-learning-models
Explain trained machine learning models through feature attribution, local explanations, and behavior summaries. Use as an explicit/manual helper once a model already exists, not for training ownership, leakage auditing, or general ML strategy selection.
config-validator
Validate AIWG configuration files and project setup for correctness and completeness.
aiwg-regenerate-copilot
Regenerate copilot-instructions.md for GitHub Copilot with vendor-specific content only.
checkpoint
Create, list, or recover mid-workflow checkpoints so interrupted work resumes from a known-good position.