undercoveragents

Use this skill when explaining shared integrations, published channels, end-user branding, and tenant-wide default model preferences.

Use this skill when explaining what Agent Alpha does, how it helps inside the admin UI, and what kind of guidance it should provide.

Use this skill when explaining what an agent is, how it is configured, and how it moves from setup to real usage.

Use this skill when explaining how tools, skills, subagents, and capabilities change what an agent can do.

Use this skill when explaining how to work inside the mission designer, shape a flow safely, and move from idea to valid workflow.

Use this skill when explaining what missions are, when to use them, and how a workflow differs from a single chat response.

Use this skill when explaining what tools are, how they differ from connectors, and which workspace owns them.

Use this skill when explaining tenants, operations, roles, login flows, and why access differs between users.

Use this skill when explaining how the admin workspace is organized, how users move between areas, and where contextual help appears.

Use this skill when explaining how admins can test agents safely, compare behavior, and review transcripts before wider rollout.

Use this skill when explaining API channels, published targets, token-based access, and how programmatic invocation differs from a client chat channel.

Use this skill when explaining what channels are, how they differ from connectors, and which shared channel types the product supports.

Use this skill when explaining branded chat channels, preview mode, default-channel behavior, and the client-facing copy controlled by a channel.

Use this skill when explaining how catalogs group skills, how builtins differ from custom catalogs, and how assignment works.

Use this skill when explaining how agent and mission test cases are structured and what kinds of expectations belong in each.

Use this skill when explaining benchmark-style agent suites, behavior assertions, and when fixtures make sense.

Use this skill when explaining how test runs are reviewed, what failed, and when a suite problem should be separated from an agent or mission problem.

Use this skill when explaining what a test suite is, how it differs from ad hoc testing, and which workspace owns it.

Use this skill when explaining where to inspect activity, verify behavior, or troubleshoot failed work across the platform.

Use this skill when the mission designer agent needs the full handbook-level guidance for workflow process, patch format, graph semantics, ports, variables, direct authoring, common patterns, and best practices.

Use this skill when explaining branches, joins, loops, variables, and the common design mistakes that stop a mission from behaving as intended.

Use this skill when explaining how missions run, how users read run history, and how to diagnose execution problems.

Use this skill when explaining how prepared retrieval knowledge supports agents and other runtime surfaces.

Use this skill when explaining what RAG is, when to use a RAG flow, and how it differs from static instructions alone.

Use this skill when explaining how a RAG flow is assembled across its stages and what users should decide during setup.

Use this skill when explaining how RAG runs are monitored, how users read progress, and how to reason about ingestion problems.

Use this skill when explaining how agents activate skills on demand and how skill resources support deeper answers.

Use this skill when explaining how skill files are packaged, imported, and bundled with reference resources.

Use this skill when explaining what skills are, why they exist, and how they differ from instructions or tools.

Use this skill when explaining how tools are attached to agents or mission steps and what users should expect at runtime.

Use this skill when explaining discovery, visibility, and the checks that make a tool ready for real use.

Use this skill when explaining how tool calls are presented to operators during live execution and in persisted history.