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 UnboundCompute/security-agent-skills --skill auditing-ssh-trust-and-agent-forwardinggit clone --depth 1 https://github.com/UnboundCompute/security-agent-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/unboundcompute/security-agent-skills/auditing-ssh-trust-and-agent-forwarding)<a href="https://agentmods.dev/skills/unboundcompute/security-agent-skills/auditing-ssh-trust-and-agent-forwarding"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/auditing-ssh-trust-and-agent-forwarding/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/unboundcompute/security-agent-skills/auditing-ssh-trust-and-agent-forwarding"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/auditing-ssh-trust-and-agent-forwarding.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 | $0.00167 | $0.02055 |
| Opus 5 | $0.00084 | $0.01027 |
| Sonnet 5 | $0.00033 | $0.00411 |
| Haiku 4.5 | $0.00017 | $0.00205 |
Grade A, and why
auditing-ssh-trust-and-agent-forwarding 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.
How it starts
The opening of the file, as written. The whole thing — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Auditing SSH trust and agent forwarding: what a connected host can turn into your identity
Secure-shell security is usually discussed as ciphers and key lengths, which are largely solved. The unsolved part is trust: who can use your authentication, whose host key you accept without checking, and whose data shapes your client's behavior. A forwarded agent lets whatever host you land on authenticate as you to everything your key opens; a proxy-command or config entry pulled from an untrusted place runs on your machine or redirects your session; a disabled host-key check turns any network position into a man in the middle; a loose authorized-key entry hands more than intended. You find these by tracing where authentication and configuration flow and asking what a compromised or untrusted endpoint can do with them.
When to use
- You are auditing how users or automation establish secure-shell trust and what it reaches.
- Agents are forwarded, host keys are accepted, or client configuration comes from shared sources.
- A compromised or multi-tenant endpoint could abuse forwarded authentication or injected config.
Scope check
Audit secure-shell trust only for hosts, accounts, and automation you own or are authorized to assess. Using a forwarded agent or a key to reach systems outside scope is not part of the audit. If you can't name the authorization, stop.
The loop
-
Map where authentication is forwarded and to which hosts. Inventory the connections and automation that forward the authentication agent, and to what hosts the socket is exposed. A forwarded agent on a single-user, trusted bastion is contained; one exposed on a multi-user, internet-facing, or lower-trust host means anyone who can reach that socket authenticates as you to everything your key opens. Determine the trust of every host the agent reaches.
-
Check who can use an exposed agent socket. Where the agent is forwarded, check the socket's permissions and the host's other users and processes. A root or co-tenant on the far host can use the socket to sign challenges as you for the duration of the session, with no access to your key itself. An agent reachable by anyone but you on an untrusted host is the finding.
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 · 141 lines · 167 tokens per session scan A 3bf8511ad00d
auditing-ssh-trust-and-agent-forwarding is a skill published in the GitHub repository UnboundCompute/security-agent-skills (5 stars, last pushed 2d ago), licensed MIT. It adds 167 tokens to every session and 2,055 once invoked, about $0.0008 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 2026-09-05.
Other skills, from other repositories
publish-threat-model
Publish a completed threat model to version control. Runs pre-flight security checks (repo visibility, secret scan), patches .gitignore with negation exceptions for the publishable files, and creates a signed git commit with threat-count metadata. Keeps pentest-tasks.yaml and all intermediate files permanently ignored.
prodcheck-review
Review this codebase against the prodcheck pre-production checklists — security, performance, scale, integrations and post-launch readiness. Use when asked to check whether a project is ready to ship, to audit an area before launch, or to work through a specific checklist. Produces evidence with file:line citations…
finishing-a-development-branch
Use when implementation is complete, all tests pass, and you need to decide how to integrate the work.
release-publish
Pre-publish and publish path for crates.io, npm installer, GitHub Releases, SBOM. Use when cutting versions, editing release.yml, or adding publish steps. Never publish from a feature branch.
requesting-code-review
Use when completing tasks, implementing major features, or before merging to verify work meets requirements.
using-git-worktrees
Use when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallback.