The Agent Skills Directory

[SAFE]: The skill consists of instructional markdown files and reference guides for the temporal and tcld command-line tools. All external resources, including domains like tmprl.cloud and temporal.io, and GitHub repositories under the temporalio organization, are official vendor resources.
[INDIRECT_PROMPT_INJECTION]: The skill provides a surface for indirect prompt injection as it instructs the agent to process untrusted data such as user-pasted error logs and SDK connection snippets to perform diagnostics.
Ingestion points: references/triage/sdk-snippet-review.md (pasted SDK connection code), references/triage/runtime-errors.md (copy-pasted error text).
Boundary markers: Absent in the input processing instructions.
Capability inventory: The skill utilizes the temporal and tcld tools, which have the capability to perform destructive actions such as deleting namespaces (references/ops/cloud-namespace-admin.md) and terminating workflows (references/triage/workflow-stuck.md).
Sanitization: Absent; the skill relies on the AI agent's internal logic and the 'Operator Discipline' requirement to propose destructive actions to the user first.
[CREDENTIALS_UNSAFE]: While the skill manages API keys and certificates, it follows security best practices. For example, references/triage/certificates.md explicitly warns: 'Never print, echo, cat, or log private-key contents.' Instructions to store secrets in environment variables or .env files are correctly handled as safe management practices.

temporal-ops