6sense
Ask an AI agent about 6sense account insights, buying signals and keyword trends.
Your team's reviewed, versioned skills library inside Claude, Claude Code and Cowork.
Opens Agent Studio, where connecting is one click. The connector URL below works in any MCP client.
Paste it into any MCP client. Setup docs
The Agentman Skills MCP server puts your team's skills library inside Claude. Members search it in plain language, apply any skill they are allowed to use, and write new skills back from the chat. It is listed in Claude's Connectors Directory, signs each member in with their own OAuth consent, and exposes 38 tools.
Built and run by Agentman
Agentman wrote this connector and hosts its endpoint.
Connection checked by Agentman on .
Listed in Anthropic’s Connectors Directory at community tier — a listing, not a partner-level review.
Anthropic states this reflects the level of review a connector received, not a security audit.
Add a skill to a skill set. Use after creating a skill set to populate it with related skills.
Cancel an in-progress chunked upload session. The upload session and any uploaded chunks are discarded.
Clone a skill (system or your own) to create a customizable copy in your workspace. The clone starts as a draft and keeps the original author attribution.
Create a new skill from SKILL.md content, after reading the creation guide and validating the content.
Create a new skill set to bundle related skills together, for a workflow, use case or project.
Delete (archive) a skill from your workspace. This is a soft delete: the skill is archived, not permanently removed.
Delete a single auxiliary resource file from a skill. On a published skill the file is removed in a new draft version; the published version is untouched until you publish.
Delete a skill set. The member skills are not deleted, only the grouping. System skill sets cannot be deleted.
Compare two versions of a skill and return a unified diff of SKILL.md and each reference file.
Search plugins by meaning using vector embeddings, for when the user describes a job rather than naming a plugin.
Semantic search for skills using natural language queries. Finds skills matching your intent, not just keywords; your workspace's skills rank before system skills.
Complete a chunked upload: assemble the chunks into the ZIP and import it as a skill.
Get aggregate statistics about available plugins: total count, plugins per category and plugins per author.
Get the official guide for creating well-structured skills: the SKILL.md format, best practices and examples.
Get a skill set's details and the skills it contains, with their prerequisites, related skills and suggested order.
Get statistics about skills in your workspace: your skills, system skills and a breakdown by category.
Import a skill from a base64-encoded ZIP file. Best for small packages; can also update an existing skill.
Import a skill from a ZIP file URL, such as a GitHub release or cloud storage. Can also update an existing skill.
Invite someone to collaborate on skills in your workspace. Only workspace admins can send invitations.
List all skill categories with counts, to discover what kinds of skills are available.
List available plugins, which bundle skills and slash commands for a role or workflow.
List the auxiliary resource files bundled with a skill: scripts, templates, reference docs and data files.
List available skill sets: bundles of related skills that work together for a domain or workflow.
List the version history of a skill, newest first, with each version's status.
List skills available in your workspace, filtered by category, tags, scope or status.
Open a plugin and see what it contains: its README, slash commands, referenced skills and expected connectors.
Open a skill and return its SKILL.md content so it can be applied to a task. Use-only skills are applied without their instructions being shown.
Open a specific, usually earlier, version of a skill and return that version's SKILL.md.
Publish a draft skill to make it available for use. Returns the published version number.
Read an auxiliary file from within a skill: a reference doc, template, example or data file.
Remove a skill from a skill set. The skill itself is not deleted.
Add or update a resource file on an existing skill. On a published skill this creates a new draft version.
Launch the Skills Explorer, an interactive view for browsing, searching and filtering skills in the chat.
Start a chunked upload session for importing a large skill archive, up to 50 MB.
Update an existing skill's SKILL.md content. On a published skill this creates a new draft version.
Update an existing skill set's name or description. System skill sets cannot be updated.
Upload a single chunk of a skill ZIP file as part of a chunked upload.
Validate and preview SKILL.md content before creating a skill, with feedback on issues. Does not create the skill.
Read from the server on 2026-10-08, including each tool's own safety annotations.
Yes. Agentman Skills has been listed in Claude's Connectors Directory since September 2026, at community tier. Community connectors have passed Anthropic's automated review but are not verified by Anthropic, so the listing is a convenience for finding and connecting it, not a security audit. Agentman wrote the connector and hosts its endpoint.
Use the directory if you work in one workspace at a time. It adds the plain connector URL, which asks you to pick a workspace when you sign in. Add a custom connector instead when you want several workspaces connected side by side, because each workspace has its own URL and Claude treats each URL as a separate connector.
No. Claude sees exactly what your own workspace login can see, because every member connects with their own OAuth consent. A skill shared with you use-only can be applied for you, but neither you nor Claude can read its instructions. Connecting an AI tool never widens your access, and an admin can revoke any member's connection.
Not permanently. delete_skill archives a skill rather than erasing it, and delete_skill_set removes only the grouping, never the skills inside it. delete_skill_resource does remove one file from a skill, but on a published skill it works on a new draft, so the published version is untouched until someone publishes the change.
tools/list from https://skills.agentman.ai/mcp through a signed-in session — 38 tools with descriptions, no safety annotations passed through, 8 October 2026Connect Agentman Skills once and your agents call these tools on their own: on a schedule, in a workflow, with nobody at the keyboard.