The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The documentation provides instructions for setting up MCP servers using standard package managers. It references the official @anthropic/mcp-atlassian package and community alternatives like mcp-atlassian (available via PyPI) and atlassian-mcp (available via NPM). These instructions are necessary for the skill's primary purpose of server integration and target well-known registries.
[INDIRECT_PROMPT_INJECTION]: The skill implements workflows that ingest data from external sources, which creates a potential surface for indirect prompt injection.
Ingestion points: Data enters the context via tool calls like jira_get_issue, jira_search, confluence_get_page, and confluence_search in references/common-workflows.md.
Boundary markers: The provided code templates do not explicitly show the use of delimiters or 'ignore instructions' warnings when processing retrieved text.
Capability inventory: The skill uses write-capable tools including jira_update_issue, jira_add_comment, confluence_create_page, and confluence_update_page across the workflow examples in references/common-workflows.md.
Sanitization: While an escapeHtml utility is provided, it is used for documentation generation; however, the skill's constraints explicitly mandate sanitizing user-provided query strings.
[CREDENTIALS_UNSAFE]: While the skill discusses the use of API tokens and OAuth secrets, it follows security best practices by discouraging hardcoding. The references/authentication-patterns.md file provides concrete implementation examples for fetching secrets from secure providers like AWS Secrets Manager, GCP Secret Manager, and HashiCorp Vault. It also includes configuration templates that use environment variables for sensitive values.

atlassian-mcp