SKILLEMALL.ai
RU Sign in

B pibox

pi-coding-agent (earendil-works) running on the network inside an aicodebox container. Exposes seven programmatic surfaces on one image — interactive shell, one-shot exec (`-p "..."`), an HTTP REST API (run/async/cancel, workspace file ops), an OpenAI-compatible `/openai/v1/chat/completions` endpoint (streaming, client-executed tool calling, response_format/JSON-schema), an MCP server at `/mcp` (mounted in API mode or as a sidecar), a Telegram bot, and a cron scheduler that fires pi on a schedule. Foreground modes (API/Telegram/Cron) are mutually exclusive except Telegram+Cron; MCP coexists with any of them. Bearer-token auth per surface (`PIBOX_API_MODE_TOKEN`, `PIBOX_MCP_MODE_TOKEN`), empty = no auth. Use when the user wants to drive pi-coding-agent programmatically over HTTP/MCP/Telegram/cron instead of a local terminal session, or needs to reason about which pibox mode/endpoint fits a given integration.

ClawHub Agent Skills author: Ciprian Mandache v0.16.1 MIT-0 3 files body ≈ 4 730 tokens open source ↗ analyzed 26 min ago

pi-coding-agent (earendil-works) running on the network inside an aicodebox container.

ProcedureTelegramDockerAI and agentstype and topics are labelled automatically from the skill text
Runs in: OpenClaw Claude / Claude Code ChatGPT / Codex Gemini CLI Cursor GitHub Copilot / VS Code Qwen Code Kimi Code CLI Z.ai (GLM) DeepSeek (Deep Code) Mistral Vibe LangChain (deepagents) Goose OpenCode
JSON
B
87/100
Overall score
Safety 60%
90
Quality 40%
82
Tests bonus
0

What is at stake

Medium-severity findings: the skill is probably honest, but read what alarmed the scanner.

Exfiltration medium severity

Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.

If you install

The instructions or scripts send data out: environment variables, keys, file contents, chat history. You may never notice, because the agent performs the upload as "part of the task".

For the author

If the upload is not needed for the task, remove it: catalogs flag such skills and delist them. If it is needed, name the destination explicitly, say what leaves the machine, and give the user a switch.

How to improve

  1. Add evals/evals.json with 4–6 real requests and expected answers: the full check will then use your cases instead of a model draft.
  2. Add a spec.yaml with triggers and assertions (skilltest init): the behaviour contract for CI.

Guard findings · 2

✓ No critical or high findings

Medium and low: 2
  • medium Exfiltration net-credential-use SKILL.md:272
    Credential used in a network call (verify the destination is the intended service)
    curl -s http://loca…080/run -H "Authorization: Bearer $SECRET" \
  • medium Exfiltration net-credential-use SKILL.md:279
    Credential used in a network call (verify the destination is the intended service)
    curl -s http://loca…080/run -H "Authorization: Bearer $SECRET" \

Files scanned: 3. Evidence is masked. Grey chips explain why severity was lowered.

Lint

  • note frontmatter-key unknown frontmatter key "homepage"

Process maturity 52/100

  • 0Result and completion. Does not say what the result is
  • 0Inputs and preconditions. Does not say what the process needs to start
  • 30Running it twice. 30 mutating operations with no state check
  • 50Failures and branches. 0 branches, has a failure section
  • 60Tools and files. Uses tools (bash, web) that frontmatter does not declare
  • 70When it triggers. States when to use, but not when not to
  • 70Execution cost. Instruction body is 4730 tokens
  • 85Steps. 18 steps, 2 vague phrases
  • 100Consistency. Name and required fields are in place
  • 100Progress reporting. Reports progress
  • medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
  • low 12 top-level sections: this looks like several domains in one skill
  • low The response is described with custom markup (6 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

  • +5Description has no quoted example phrases that should trigger the skill
  • +4Description does not say when NOT to use the skill (false activations)
  • +3Description length 920: 120–800 characters recommended
  • +3Output format is not stated: the model decides each time
  • -2localhost URLs: will not work for another user
  • +1No license
  • +2Single-language instructions
  • +4Structure: 13 headings
  • +3Step-by-step instructions: 18 items
  • +4Has examples (14 code blocks)
  • +4Reference files are cited in the instructions (1 of 1)

Quality base 70; lint remarks subtract, signals add up to 100. Result: 82.

External checks

ClawHub: suspicious
Review before installing: this skill documents a networked coding-agent container that can run prompts, modify or delete workspace files, and is commonly launched with broad networking and an unpinned Docker image.
LLM: suspicious (medium) · 9 Sept 2026