Observability Designer
v2.1.1Observability Designer (POWERFUL)
⭐ 0· 815·6 current·6 all-time
byAlireza Rezvani@alirezarezvani
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
Name and description match the included artifacts: three Python scripts (SLO designer, alert optimizer, dashboard generator), sample assets, references, and README. There are no unrelated required binaries, environment variables, or config paths declared; the requested capabilities align with designing observability artifacts and generating configs for Prometheus/Grafana/Alertmanager.
Instruction Scope
SKILL.md and README focus on generating SLO frameworks, optimizing alerts, and producing dashboard JSON. Examples reference running the provided Python scripts with input files (assets or user-provided service definitions). Runbook examples contain common operational commands (curl, kubectl) as troubleshooting guidance — these are examples and not instructions to exfiltrate secrets. Nothing in the instructions asks the agent to read unrelated host files or to harvest credentials.
Install Mechanism
There is no install spec — this is instruction-only from the platform perspective. The bundle contains Python scripts and sample assets; the README states only Python 3.7+ and standard library are required. Because no external downloads or package installers are declared, the install surface is low, but executing the scripts will write/read files locally when you run them.
Credentials
The skill declares no required environment variables or credentials. Sample configurations include placeholder-looking values (PagerDuty routing_key, Slack webhook URL) and example runbook commands; those are expected for alerting examples but will be sensitive if you replace them with real secrets. If you run these scripts with real production configs, they may read or output secrets in generated configs or reports — the skill itself doesn't demand credentials up front.
Persistence & Privilege
always:false and disable-model-invocation:false (normal). The skill does not request permanent platform presence or modify other skills. Execution appears user-driven (run python scripts), so it has no elevated persistence or platform-wide privileges.
Assessment
This package appears coherent for designing observability artifacts, but exercise caution before running any included scripts against production data or with real credentials: 1) Review the full Python scripts locally (they were present in the bundle) to confirm they don't make unexpected network calls or upload data; 2) Run them in an isolated environment (container or VM) or a Python virtualenv first; 3) Replace or redact example webhook/routing keys in sample configs before importing into live Alertmanager/PagerDuty/Slack; 4) If you need higher assurance, provide the full, untruncated source of the three scripts for a line-by-line review — I flagged medium confidence because I could not verify every line of the scripts in this review.Like a lobster shell, security has layers — review code before you run it.
latestvk9784e6ppbyzhs4q1hzcqteg9n82mvzn
License
MIT-0
Free to use, modify, and redistribute. No attribution required.