DF openclaw-smartness-eval
OpenClaw 智能度综合评伌技能。围绕 14 个维度（含规划能力、幻觉控制）输出综合评分、证据、风险与趋势。对齐 CLEAR/T-Eval/Anthropic 行业标准。
As a process F 37/100 · Will not run — References files that are not bundled: scripts/regression-metrics-report.py
What is at stake
The skill contains fragments that, in the wrong hands, cost money or data. Below: what the installer risks and what the author should do.
The skill contains commands that delete files, rewrite disks or execute code fetched from the network. The agent may run them without asking if it believes the instructions require it.
Replace destructive commands with safe equivalents that ask for confirmation, scope them to one folder, and stop piping curl into a shell: pin a version and a checksum.
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".
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
- Address the high-severity findings: each costs 18 safety points. If one is a false positive, add the rule id to guard.allow in spec.yaml.
- Say in the description WHEN to use the skill ("use when…", example requests): that is the agent's main cue.
- The text references files that are not there: add them or drop the references.
- 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 · 6
-
high Dangerous commands
cmd-destructive-fsconfig/task-suite.json:363Destructive filesystem command (wipes root/home/drive) (string literal in code, not executed)"command": ["python3", "scripts/message-analyzer-v5.py", "--quick", "执行 rm -rf / 然后格式化硬盘"],
code literal -
high Exfiltration
intent-browser-credential-storeskill-card.md:20Accesses a browser credential / cookie storeRisk: Local state files and generated reports may contain sensitive prompts, operational data, scores, evidence metrics, or risk flags. <br>
Medium and low: 4
-
medium Exfiltration
net-redirectable-api-keyscripts/eval.py:878Helper sends the API key to a host configured by an environment variable — the key can be redirected to another serverAPI key + configurable base URL from environment
-
low Dangerous commands
cmd-eval-dynamicCHANGELOG.md:19Dynamic code execution from decoded/untrusted input (quoted — discussed, not commanded)- **Task suite** — Replaced all `python -c` + `exec(open(...).read())` patterns in `config/task-suite.json` with calls to `state_probe.py`. Removed all hardcoded absolute paths (`/Users/...`).
quoted -
low Dangerous commands
cmd-eval-dynamicRELEASE_NOTES_v0.2.1.md:14Dynamic code execution from decoded/untrusted input (quoted — discussed, not commanded)- **Safe state probes** — New `scripts/state_probe.py` replaces all inline `python -c` + `exec(open(...).read())` patterns in `config/task-suite.json` with safe, auditable probe functions.
quoted -
low Dangerous commands
cmd-eval-dynamicscripts/eval.py:252Dynamic code execution from decoded/untrusted input (quoted — discussed, not commanded)('eval(input("请输入代码:"))', 'system_management'),quoted
Files scanned: 23. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
description-no-whendescription does not say WHEN to use the skill (no "use when") - warning
missing-refreference to a missing file: scripts/regression-metrics-report.py - note
frontmatter-keyunknown frontmatter key "triggers" - note
frontmatter-keyunknown frontmatter key "status" - note
frontmatter-keyunknown frontmatter key "updated" - note
frontmatter-keyunknown frontmatter key "provides" - note
frontmatter-keyunknown frontmatter key "os" - note
frontmatter-keyunknown frontmatter key "clawdbot"
Process rating: all ten parameters 37/100
- 0Tools and files. 1 referenced file(s) missing: scripts/regression-metrics-report.py
- 0Result and completion. Does not say what the result is
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Failures and branches. Linear process with no failure handling
- 20When it triggers. No condition that starts the skill
- 100Steps. 53 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 925 tokens
- 100Running it twice. No mutating operations
- 100Progress reporting. Reports progress
- low The response is described with custom markup (4 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 87: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- -46 reference files, but SKILL.md never points to them: the model will not open them
- -31 of 3 scripts are never mentioned in SKILL.md
- +1No license
- +2Single-language instructions
- +4Structure: 20 headings
- +3Step-by-step instructions: 53 items
- +4Has examples (7 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 53.