Skills + Tactet

Build each repeatable task on your company's systems as an agent in Tactet, and install one skill in the agent harness. The skill makes two calls on your agent server: search finds the agent for a request, and execAgent runs it. The agent harness keeps the agentic loop. The work on your systems runs on the agent server.

YOUR ORGANIZATIONCLAUDE CODEClaude CodeTactet skillsearch · execAgentTactet Agent ServerAgent runtimeagents run hererequestanswerYour systemsAPIs · databases · MCPLLMsany vendor, per profile
A skill in Claude Code asks Tactet; the systems and models stay behind Tactet

The usual alternative is a skill per workflow, with MCP servers or scripts that reach company systems from each laptop. There, Claude reads the instructions and writes and runs the steps itself, on every run, in every person's session.

The table lists what changes when those workflows become Tactet agents. What Tactet adds describes each change, with its sources.

At a glance

What changesA skill per workflow, with MCP servers or scriptsOne skill, agents in Tactet
Where a workflow livesIn instructions Claude interprets on each runIn a fixed, validated, versioned agent that runs on your server
What enters Claude's contextThe instructions, every step Claude writes, every intermediate resultThe skill's instructions, the names and titles of matching agents, each agent's output
Model tokens per runFinding the skill, loading it, and a model call for each step Claude writesOne skill and two calls by Claude, plus the agent's llm steps; search and other commands cost none
What the model work costsClaude usage for every stepClaude usage for the two calls; each llm step at the price of its model, or no per-token charge on a model you host
When the assistant's model changesEach skill is due a retestOne skill is due a retest
Models a step can useClaude modelsAny vendor's, or one you host with Ollama, vLLM or SGLang, chosen per step in a profile
What a laptop holdsA route and a credential for each system, or for the MCP server in front of itOne revocable token for the agent server
What can act on your systemsWhat the model decides at run time, within the person's access and permission rulesPublished agents, inside their profiles, allowlist, time and cost limits
Record of a runSession telemetry, if you collect itAgent, version, person, duration, result and cost, in built-in reports
Adding a workflowA skill has to reach each person's machine or accountAn agent is published

What stays with Claude

The agent harness keeps the agentic loop. It splits a request into parts, sends each through search and execAgent, decides the order and the conditions from the results, and composes one answer.

A support engineer types: Has order 48812 shipped? If not, open a ticket with the warehouse. Claude runs the order-lookup agent, reads "not shipped", then finds and runs the warehouse-ticket agent in the same conversation. Tactet did each step inside the company. Claude decided the order and the condition.

Claude also keeps everything it does without Tactet: files, code, and requests nobody has built an agent for. When nothing fits, search returns no match instead of guessing. In published tests it refused 100% of unrelated questions and 82 to 86% of in-domain questions that no agent covered. Claude then answers as it normally would.