The Agent Skills Directory

[DATA_EXFILTRATION]: The skill provides automation scripts in references/api-monitoring.md that access the sensitive local Terraform credentials file at ~/.terraform.d/credentials.tfrc.json to extract API tokens for authentication with HashiCorp's official platform.
[COMMAND_EXECUTION]: The instructions direct the agent to execute various terraform stacks CLI commands for infrastructure management and provide shell scripts utilizing curl and jq to interact with external APIs.
[EXTERNAL_DOWNLOADS]: The skill configuration and documentation reference the use of Terraform providers and modules sourced from public registries and private registries hosted on the app.terraform.io domain.
[PROMPT_INJECTION]: The skill exhibits metadata poisoning and attack surface vulnerabilities for indirect injection:
The YAML frontmatter in SKILL.md contains a copyright notice attributed to "IBM Corp", which contradicts the designated skill author "Lua2147". This deceptive metadata may lead users or systems to misattribute the skill's origin and safety.
The skill is designed to process and act upon user-supplied HCL configuration files, which represents an attack surface for indirect prompt injection.
Ingestion points: Terraform configuration files (.tfcomponent.hcl, .tfdeploy.hcl) located at the repository root.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the processing logic.
Capability inventory: The skill enables shell command execution, network requests, and access to local credential files.
Sanitization: The skill relies on the external terraform binary for syntax and schema validation of processed files.

terraform-stacks