BB tunnel-doctor
Diagnoses concrete tunnel and proxy-path failures across macOS and Windows/WSL: Tailscale routing, proxy env/system bypass, SSH double tunneling, VM/container propagation, stalled DNS, TUN DIRECT split-brain, Windows-host TUN cascades, and single-hop or chained proxy node/exit throughput. Use when Tailscale ping works but SSH/HTTP fails, browser returns 503 while curl works, Git SSH closes through the proxy, Docker pull/build fails behind TUN, getaddrinfo stalls while nslookup is fast, raw probes report physically impossible results, domestic DIRECT-routed sites fail while proxied sites work, a blocked port could be your tunnel or the destination's firewall, or the proxy is reachable but real downloads and full Git clones crawl. Also use when Git reports "failed to begin relaying via HTTP", ssh -vvv freezes at "debug2: resolving", ping works but dig times out, or setting up Tailscale SSH to WSL. Use debugging-network-issues only when the root cause remains unknown or belongs to an application/protocol layer.
Diagnoses concrete tunnel and proxy-path failures across macOS and Windows/WSL: Tailscale routing, proxy env/system bypass, SSH double tunneling, VM/container…
As a process B 69/100 · Nearly there — weak spots: result and completion, execution cost
What is at stake
Medium-severity findings: the skill is probably honest, but read what alarmed the scanner.
Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.
The skill asks for more than the task needs: broad tool access, credential environment variables, binaries. Every extra permission widens the damage from a mistake or a compromise.
Narrow allowed-tools and the variable list to the minimum; replace binaries with readable sources or scripts.
How to improve
- The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 3
✓ No critical or high findings
Medium and low: 3
-
medium Broad scope
meta-broad-allowed-toolsSKILL.md:1Broad tool permissions pre-approved: Bashallowed-tools: Read Grep Edit Bash
-
low Dangerous commands
cmd-pipe-to-shell-known-hostSKILL.md:1043Pipe-to-shell installer from a well-known host (still executes remote code) (documentation of a security skill)curl -fsSL https://tailscale.com/install.sh | sh
security skill -
low Exfiltration
read-dotenvSKILL.md:1271Reads a .env file (documentation of a security skill)# 2. Copy .env from local machine (run on local)
security skill
Files scanned: 7. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
body-longSKILL.md body ≈ 20904 tokens (recommended < 5000); move details to references/
Process rating: all ten parameters 69/100
- 0Result and completion. Does not say what the result is
- 10Execution cost. Instruction body is 20904 tokens: crowds the task out of the window
- 50When it triggers. No condition that starts the skill
- 70Inputs and preconditions. Inputs and preconditions are listed
- 85Steps. 124 steps, 2 vague phrases
- 100Tools and files. Tools declared in frontmatter
- 100Failures and branches. 2 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Running it twice. Mutating operations check current state
- 100Progress reporting. Reports progress
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- low The response is described with custom markup (32 tags): a typed call is more reliable
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +4Description does not say when NOT to use the skill (false activations)
- +3Description length 1023: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- -2localhost URLs: will not work for another user
- -215 emoji in the instructions: noise for the model
- +1No license
- +2Single-language instructions
- +5Description quotes 2 example trigger phrases
- +4Structure: 50 headings
- +3Step-by-step instructions: 124 items
- +4Has examples (71 code blocks)
- +4Reference files are cited in the instructions (5 of 5)
- +3All 1 scripts are documented
Quality base 70; lint remarks subtract, signals add up to 100. Result: 79.