skills/agentfs-cli/SKILL.md
Inspect filesystem state with the local `agentfs` CLI. Use when an agent needs a bounded directory snapshot, file previews, or optional hashes without traversing an uncontrolled tree.
npx skillsauth add sebastianboehler/agent-cli-utils agentfs-cliInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Use agentfs to inspect a file tree with explicit limits so the result stays cheap enough for low-compute systems and downstream agent parsing.
Prefer the installed binary when it exists:
agentfs -root . -max-depth 2 -format json
If working inside this repository or the binary is not installed, run:
go run ./cmd/agentfs -- -root . -max-depth 2 -format json
-root to set the inspection root.-max-depth to keep traversal bounded.-max-entries to avoid oversized snapshots.-preview-lines for cheap text previews.-hash only when content integrity matters.-hidden only when hidden files are relevant.-format json or -format yaml for automation, text for review.Take a bounded repo snapshot:
go run ./cmd/agentfs -- -root . -max-depth 2 -preview-lines 2
Inspect logs in text form:
go run ./cmd/agentfs -- -root /var/log -max-depth 1 -preview-lines 3 -format text
tools
Use the local `company` CLI for German Handelsregister, OffeneRegister, and OpenCorporates company lookup with structured output. Prefer it when an agent needs registry-backed company research before lead scoring, outreach, or CRM/database updates.
tools
Send or validate SMTP mail flows with the `agentsmtp` CLI. Use when an agent needs a small SMTP submission tool with provider presets for Gmail or Google Workspace, structured output, and configurable auth through password or app-password.
tools
Use the local `agentrunpod` CLI to submit, inspect, and cancel RunPod serverless jobs with structured output. Prefer it when an agent needs an env-configurable wrapper around the RunPod HTTP API instead of ad hoc curl commands.
tools
Execute shell commands with the local `agentrun` CLI. Use when an agent needs timeout control, bounded stdout and stderr capture, or structured command results instead of direct raw process execution.