CF mjolnir-brain
AI Agent 自进化记忆系统。核心功能 100% 本地运行 (纯 Markdown+JSON)。网络备份/节点通信为可选 opt-in，需手动配置凭证。
As a process F 35/100 · Will not run — References files that are not bundled: scripts/*.sh
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 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.
Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.
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.
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
- 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 · 11
✓ No critical or high findings
Medium and low: 11
-
medium Dangerous commands
cmd-persistenceINSTALL.md:165Persistence mechanism (cron / launchd / scheduled task / autorun registry) (detector / deny-list definition)crontab -l | grep -v openclaw | crontab -
detector -
medium Dangerous commands
cmd-shell-rcplaybooks/multi-user-setup.md:136Writes to a shell startup fileecho 'export MJOLNIR_USER=<their_username>' >> ~/.bashrc
-
medium Exfiltration
net-credential-usescripts/workspace_backup.sh:83Credential used in a network call (verify the destination is the intended service)curl -s -X MKCOL -u "$BACKUP_WEBDAV_USER:$BACKUP_WEBDAV_PASS" "$BACKUP_WEBDAV_URL" 2>/dev/null || true
-
medium Broad scope
meta-agent-memory-dumptemplates/HEARTBEAT.mdAgent memory / workspace files bundled with the skill (5) — likely a workspace dump with personal data or tokenstemplates/HEARTBEAT.md, templates/IDENTITY.md, templates/MEMORY.md, templates/SOUL.md, templates/USER.md
-
low Dangerous commands
cmd-cron-mentionINSTALL.md:91Mentions editing / listing crontabcrontab -e
-
low Dangerous commands
cmd-cron-mentionINSTALL.md:149Mentions editing / listing crontabcrontab -l | grep openclaw
-
low Dangerous commands
cmd-cron-mentionINSTALL.md:165Mentions editing / listing crontabcrontab -l | grep -v openclaw | crontab -
-
low Dangerous commands
cmd-cron-mentionREADME.md:94Mentions editing / listing crontab# crontab -e
-
low Dangerous commands
cmd-cron-mentionSKILL.md:98Mentions editing / listing crontab# crontab -e 添加定时任务
-
low Dangerous commands
cmd-install-from-urlstrategies.json:74Installs a package from an untrusted URL / archive (string literal in code, not executed; security demo / example){"action": "Use mirror: pip install -i https://mirrors.example.com/pypi/simple/", "successRate": 0.90, "tries": 10, "env": "behind firewall"},code literaldemo
A further 1 matches are quotations in this security skill's documentation and are not counted as findings.
Files scanned: 51. 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/*.sh
Process rating: all ten parameters 35/100
- 0Tools and files. 1 referenced file(s) missing: scripts/*.sh
- 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
- 0Progress reporting. Says nothing while it works
- 20When it triggers. No condition that starts the skill
- 100Steps. 13 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 423 tokens
- 100Running it twice. No mutating operations
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 78: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- -214 emoji in the instructions: noise for the model
- -313 of 13 scripts are never mentioned in SKILL.md
- +1No license
- +2Single-language instructions
- +4Structure: 12 headings
- +3Step-by-step instructions: 13 items
- +4Has examples (4 code blocks)
- +4Reference files are cited in the instructions (4 of 8)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 62.