agentmint-hermes-runner 0.8.0

Route Hermes delegate_task(background=True) to named, persistent AgentMint subagents.

agentmint-hermes-runner

Python adapter that bridges Hermes' delegate_task(background=True) to named, persistent AgentMint subagents — specialists that accumulate /workspace/MEMORY.md across calls. Two setup paths; combine freely:

Mode	What it does	Best for
Persistent	delegate_task(background=True) routes to a pre-minted named subagent (via default_agent_name, or via an agentmint-<name> entry the LLM passes in toolsets). Its /workspace/MEMORY.md accumulates across calls.	One long-running specialist that learns OR LLM picking among many specialists per call (via toolset hack — see below)
Plugin tool	Hermes auto-discovers a new agentmint_delegate(agent_name, goal, ...) tool from this package's entry-point. LLM picks the subagent per call.	Cleaner per-call routing when you can expose a new tool to the LLM

The Hermes-installable skill that drives this adapter lives in a separate catalog repo: mesutcelik/agentmint-skills — hermes skills install mesutcelik/agentmint-skills/hermes-delegate-task. The skill references this package by its PyPI name (pip install agentmint-hermes-runner).

Status

v0.8.0 — alpha. Auth backends: BearerAuth (Stripe-Link), TempoAuth (Tempo USDC.e — Tier 1 direct only; the delegate_task patches require Bearer). Requires AgentMint API ≥ 0.10.0 for workspace_files support; ≥ 0.7.0 for the rest.

Setup — persistent (default routing)

Every delegate_task(background=True) lands in one named subagent:

import os
from agentmint_hermes_runner import AgentMintDispatcher, BearerAuth, install_delegate_task_wrapper

dispatcher = AgentMintDispatcher(auth=BearerAuth(jwt=os.environ["AGENTMINT_JWT"]))
install_delegate_task_wrapper(dispatcher, default_agent_name="default-worker")

Pre-mint default-worker once (agent.create via curl). Its MEMORY.md grows across every delegation.

Setup — persistent (per-call routing via toolset hack)

When the LLM should pick a specialist per call, but you don't want to expose a new tool, encode the target name in toolsets:

# Operator wires the adapter the same way — default_agent_name is optional:
install_delegate_task_wrapper(dispatcher, default_agent_name="default-worker")

# The LLM then dispatches like this:
delegate_task(
    background=True,
    goal="Review the diff",
    toolsets=["terminal", "file", "agentmint-pr-reviewer-myrepo"],
)

The adapter parses agentmint-pr-reviewer-myrepo from toolsets, routes to that subagent (overriding default_agent_name), and strips the entry from the toolset list before composing the prompt.

This is a workaround for Hermes' delegate_task not accepting a dispatcher-target argument. A formal proposal is in docs/hermes-feature-request.md — when an upstream extension lands, this hack will be deprecated.

Setup — plugin tool (named fleet, cleanest)

from agentmint_hermes_runner import AgentMintDispatcher, BearerAuth, set_dispatcher

dispatcher = AgentMintDispatcher(auth=BearerAuth(jwt=os.environ["AGENTMINT_JWT"]))
set_dispatcher(dispatcher)   # registers agentmint_delegate via the entry-point

Hermes' plugin discovery (hermes_agent.plugins entry-point) auto-registers agentmint_delegate when this package is pip-installed. The LLM can now call:

agentmint_delegate(agent_name="reviewer-myrepo", goal="Review the diff", async_=True)
agentmint_delegate(agent_name="support-acme",    goal="Reply to ticket #42", async_=True)

Combine with install_delegate_task_wrapper(dispatcher, default_agent_name=...) if you want BOTH delegate_task and agentmint_delegate to dispatch to AgentMint.

See examples/persistent.py and examples/plugin.py for complete operator setup snippets.

Install

pip install agentmint-hermes-runner

Test

pip install -e ".[dev]"
pytest
ruff check .

Lower-level surface

If you want to drive AgentMint directly without the delegate_task patch or the plugin tool:

result = dispatcher.dispatch(
    agent_name="reviewer-myrepo",
    goal="Review the diff at /workspace/pr-42.diff and flag risks.",
    context="Project at /workspace, Python 3.11, uses Flask + PyJWT.",
    toolsets=["terminal", "file"],     # "web" raises UnsupportedToolset
    role="leaf",                        # or "orchestrator"
    max_iterations=50,
    child_timeout_seconds=600,
    workspace_files=[                   # ship inputs into the sandbox before the run
        {"path": "/workspace/pr-42.diff", "content": "diff --git a/foo ..."},
    ],
    cleanup_paths=["/workspace/pr-42.diff"],  # wipe them after the run
)

# Batch dispatch (Hermes tasks=[…] analog):
results = dispatcher.dispatch_batch(
    tasks=[
        Task(agent_name="researcher-wasm", goal="WASM 2026 survey"),
        Task(agent_name="researcher-riscv", goal="RISC-V 2026 survey"),
    ],
    max_concurrent_children=3,
    child_timeout_seconds=900,
)

Known unsupported

• toolsets=["web"] — no canonical AgentMint web-fetch skill yet. Raises UnsupportedToolset.
• max_spawn_depth — AgentMint sandboxes aren't structurally bounded by depth.
• Tempo + the delegate_task patches — polling against agent.run.status is Bearer-only. Tempo customers can use Tier 1 (direct curl) but not the install/plugin paths above.

License

MIT