The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to have the AI ingest and process log files which are external data sources created by other system processes. 1. Ingestion points: The files located at ~/.sf/vscode-spans/*.jsonl and ~/.sf/vscode-appinsights/*.jsonl serve as untrusted data inputs. 2. Boundary markers: There are no instructions to use delimiters or ignore embedded instructions when the agent reads these log bodies. 3. Capability inventory: The skill utilizes shell capabilities including ls, jq, sort, and rm for file management and parsing. 4. Sanitization: The instructions do not specify any sanitization or validation of the log content before the AI processes it.
[DATA_EXFILTRATION]: The skill accesses local file paths containing sensitive information about the developer's environment. Access: It reads from ~/.sf/vscode-spans/ and ~/.sf/vscode-appinsights/, which contain identifiers such as userId, orgId, cliId, and webUserId associated with Salesforce organizations. Context: This access is central to the skill's primary purpose for debugging and local observability.
[COMMAND_EXECUTION]: The skill provides various shell commands for the agent or user to execute locally. Commands: Includes ls for locating the latest logs, jq for querying trace data, rm for clearing telemetry files, and npm run for starting a local span server component.

span-file-export