The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the lgtm-cli from the vendor's GitHub repository (github.com/pokgak/lgtm-cli) using the uv tool manager. This is a documented part of the skill's setup process.
[COMMAND_EXECUTION]: The skill executes local commands using the lgtm CLI and kubectl to retrieve observability data and establish port-forwarding tunnels to internal services.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core function of ingesting and summarizing log data which may contain attacker-controlled strings.
Ingestion points: Log content is retrieved from Loki via the lgtm loki query command in SKILL.md.
Boundary markers: The skill lacks explicit delimiters or instructions to ignore embedded commands when log data is passed to subagents for summarization.
Capability inventory: The agent possesses Bash access and can execute further network queries, trace searches, and configuration commands.
Sanitization: There is no evidence of sanitization or escaping of log content before it is processed by the AI models.

lgtm