Skip to main content
Glama
Dogtiti

dsh-plugin-subagent-bridge

by Dogtiti

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
portNoLoopback listening port; a fixed port keeps the Codex-side URL valid across restarts.47821
enabledNoMaster switch; when disabled the plugin does not listen on a port and does not write the endpoint file.true
maxHopsNoMaximum delegation depth across tools (1 = only one layer allowed, matching the default maxDepth: 1 of Codex / DSH).1
codexBinNoCodex executable; when it contains '/', it is resolved strictly as a path and an error is raised instead of falling back.codex
maxWorkersNoGlobal fallback cap to prevent many projects from saturating the machine together.12
workerTtlMsNoRecords not updated within this time stop occupying a slot (expired slots are released automatically).86400000
workersFileNoLocation where the worker registry and time baselines are persisted (0600).~/.dsh/bridge/workers.json
codexSandboxNoDefault sandbox for codex_dispatch (has no effect when resuming).workspace-write
endpointFileNoLocation where the endpoint URL and token are persisted (0600).~/.dsh/bridge/endpoint.json
waitTimeoutMsNoDefault timeout for dsh_wait (1s to 240s; clients may override with timeoutMs).240000
codexTimeoutMsNoTimeout for a single codex_dispatch; on timeout SIGTERM is sent first, then SIGKILL.300000
maxResultCharsNoMaximum number of characters returned by a single dsh_read_result call (excess keeps the tail).20000
readWaitSecondsNoDefault long-poll duration for dsh_read_result (max 55; clients may override with waitSeconds).25
dispatchWindowMsNoLength of the rate-limit window in milliseconds.60000
dispatchPerMinuteNoNumber of dispatches allowed within the rolling window (spawn / task / dispatch / followup / codex_dispatch combined).10
maxWorkersPerRootNoWorker cap per working directory (cwd is bucketed automatically; callers do not pass an owner).4
taskTimeoutSecondsNoDefault limit for dsh_task waiting for one round to finish (clients may override with timeoutSeconds, max 600).300

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

Capabilities

Server capabilities have not been inspected yet.

Tools

Functions exposed to the LLM to take actions

NameDescription

No tools

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

Maintenance

ActivityMaintained
ResponsivenessUnresponsive