Skip to main content
Glama
arctop

Seamless

by arctop
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Go
Local

session_start

Idempotent

Bind an agent connection to a work session and return its project briefing. Use it for hookless clients, switching projects, or resuming a named session.

Instructions

Bind this connection to a work session and return its project briefing. Claude Code and Codex agents running the Seamless hooks are bound to their own session automatically (the 'Seam session' line in the briefing) and do not need this. Call it to bind a client without hooks, to move this connection to another project, or to resume a named session. Later calls inherit the session's project, so they need no project argument.

Input Schema

NameRequiredDescriptionDefault
cwdNoAbsolute working directory. The git repository it is in decides the project, and a repository maps itself to a project on its first session -- no setup step (`seamlessd map-repo` only overrides the derived slug)
hostNoMachine this agent runs on (hostname). Defaults to the X-Seamless-Host header, then to the daemon's own host; pass it only when dialling a daemon on another machine
nameNoResume the session with this name: a sess/* name from an earlier call, or the cc/... or cx/... on your briefing's 'Seam session' line. A new name starts a separate session under that name; omit it to use your own session, or a fresh one
modelNoModel id powering this agent, exactly as the provider names it (e.g. claude-fable-5, gpt-5.5). Stamped onto memories/notes this session writes; hooks keep it current for Claude Code/Codex sessions, so pass it mainly from other clients
projectNoProject slug to bind. Defaults to the project of the git repository cwd is in, then to your own session's project. If cwd is in a repository too, the two must agree. project=global binds the global scope; an unknown slug creates that project.
repo_rootNoAbsolute git repository root enclosing cwd, resolved on YOUR machine. Required when host is not the daemon's: the daemon cannot read your filesystem to find it
repo_originNoThe repository's origin remote URL, which is how the same repo checked out on two machines is recognized as one project
main_worktree_rootNoAbsolute root of the repository's MAIN checkout when repo_root is a linked worktree; defaults to repo_root

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed4 schema fields changedv0.7.4
    • changedInput schema / properties / cwd / description
      Previous value: -"Absolute working directory; auto-mapped to a project from the repo root on a repo's first session (no setup step -- `seamlessd map-repo` only overrides the derived slug)"New value: +"Absolute working directory. The git repository it is in decides the project, and a repository maps itself to a project on its first session -- no setup step (`seamlessd map-repo` only overrides the derived slug)"
    • changedInput schema / properties / name / description
      Previous value: -"Optional stable session name; reusing a name resumes that session"New value: +"Resume the session with this name: a sess/* name from an earlier call, or the cc/... or cx/... on your briefing's 'Seam session' line. A new name starts a separate session under that name; omit it to use your own session, or a fresh one"
    • addedInput schema / properties / project
      Added value: +{
+  "description": "Project slug to bind. Defaults to the project of the git repository cwd is in, then to your own session's project. If cwd is in a repository too, the two must agree. project=global binds the global scope; an unknown slug creates that project.",
+  "type": "string"
+}
    • removedInput schema / properties / source
      Removed value: -{
-  "description": "what began this session (default explicit)",
-  "enum": [
-    "startup",
-    "resume",
-    "clear",
-    "compact",
-    "explicit"
-  ],
-  "type": "string"
-}
  2. Changed4 schema fields changedv0.5.1
    • addedInput schema / properties / host
      Added value: +{
+  "description": "Machine this agent runs on (hostname). Defaults to the X-Seamless-Host header, then to the daemon's own host; pass it only when dialling a daemon on another machine",
+  "type": "string"
+}
    • addedInput schema / properties / main_worktree_root
      Added value: +{
+  "description": "Absolute root of the repository's MAIN checkout when repo_root is a linked worktree; defaults to repo_root",
+  "type": "string"
+}
    • addedInput schema / properties / repo_origin
      Added value: +{
+  "description": "The repository's origin remote URL, which is how the same repo checked out on two machines is recognized as one project",
+  "type": "string"
+}
    • addedInput schema / properties / repo_root
      Added value: +{
+  "description": "Absolute git repository root enclosing cwd, resolved on YOUR machine. Required when host is not the daemon's: the daemon cannot read your filesystem to find it",
+  "type": "string"
+}
  3. Changed1 schema field changedv0.1.3
    • changedInput schema / properties / source / description
      Previous value: -"what began this session (default startup)"New value: +"what began this session (default explicit)"
  4. First observedv0.1.0

TDQS

A4.1/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations declare the safety profile (mutation, idempotent, non-destructive), so the burden is lighter. The description adds genuinely non-annotation behavioral context: the call returns a briefing, and 'later calls inherit the session's project, so they need no project argument' — a stateful, connection-scoped side effect an agent must know.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Four sentences, each front-loading purpose before conditions. The first sentence leads with the core action; the exception and usage clauses follow. No filler sentences, though it is on the longer side for the amount of new information.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

No output schema exists, so the mention that the call returns 'its project briefing' usefully fills that gap. For an 8-parameter, non-destructive mutation tool with annotations and full schema coverage, the description covers purpose, usage conditions, and stateful behavior — minor gaps only.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema already documents all 8 parameters in detail. The description only loosely gestures at parameter meaning ('resume a named session', 'move this connection to another project') and adds no syntax or format not already in the schema. Baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource: 'Bind this connection to a work session and return its project briefing.' It also clarifies the automatic-hook case that does not need this call. It stops short of explicitly differentiating from sibling tools like session_update or session_end, but the core action is unambiguous.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives explicit when-to-use conditions ('bind a client without hooks, to move this connection to another project, or to resume a named session') and an explicit when-NOT-needed case ('Claude Code and Codex agents running the Seamless hooks ... do not need this'). Nothing about invocation timing is left to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server

Other Tools