graph
v1.0.9by BANG404
Split work into dependent tasks, run independent tasks in parallel, and view progress and results in a task graph.
Package we checked
plugin.json sits at the package root in BANG404/openagent-graph, which is the plugin directory the 1.0.0 spec describes. Its $schema is https://agent-plugins.org/schemas/1.0.0/plugin.schema.json. The manifest lists 1 skill (graph) and 1 MCP server (graph). Extension keys: openagent. Stored mcp.json transports: graph stdio.
What graph does, in the publisher's words
The standard Agent Plugin package for dependency aware Graph execution. The package owns the graph schema, reducer, persistence, child conversations, parallel scheduling, prompts, recovery, MCP tools, and progress projection. OpenAgent supplies the same generic capability bridge available to every plugin: conversations, branches, Agent submission and wake, roles, events, cancellation, permissions, and ordinary checkpoint storage.
The package uses the portable Agent Plugins 1.0.0 format and is an ordinary package; the Runtime does not select or implement Graph behavior. The graph command is exposed as /graph through the generic package-id alias; /graph:graph remains the full namespaced route. mcp.json starts bin/graph-mcp.mjs, which persists Graph state under PLUGIN_DATA/graphs/ and wakes child Agents through agent.wake.
The slash command only supplies the planning/execution prompt. It does not write package state or publish an empty running projection. A Graph starts when the Agent successfully calls create_goal_graph; that tool persists the DAG and publishes its first projection before scheduling nodes.
From the project README, punctuation lightly normalized · Open the README on GitHub
- Skills
- 1
- MCP servers
- 1
- Stars
- 0
- License
- MIT
- Repo created
- 2026-09-28
- Last pushed
- 2026-10-17
- Publisher type
- User
- Version
- 1.0.9
Links
Skills · 1
- graph
- Use OpenAgent Graph Mode when a task can be represented as dependent nodes that should execute with durable progress and parallel scheduling.
Descriptions come from the frontmatter of each SKILL.md, punctuation lightly normalized.
MCP servers · 1
- graphstdio
- node ${PLUGIN_ROOT}/bin/graph-mcp.mjs
Read from the plugin's own mcp.json. We list environment variable names, never values.
Category
Other. Assigned by this directory. Agent Plugins 1.0.0 has no category field, so no manifest declares one.
Keywords
openagent · graph · goal · checkpoint
Related by publisher and keywords
Create OpenAgent plugins from requirements or templates, then develop, test, and validate them for installation.
A warm taupe theme with rose accents, cyan charts, and 14px text; supports light and dark modes.
Find symbols, follow references, and edit code by symbol with Serena in the active project.
Telegram messaging connected to OpenAgent conversations.
Long-running task patterns distilled from OpenAI Codex harness v0.149.0: tool output budgeting, context pressure compaction, parallel sub-agent fan-out, structured plan streaming, self-review, sub-agent delegation, world-state tracking, background task management, thread-level goal persistence, per-sub-task model routing, completion auditing, fork-context decision, sub-agent family tracking, goal token budgeting, error recovery strategy, retry with backoff, streaming output reader, session handoff, long-term memory, skill auto-selection, plugin authoring helper, tool discovery pattern, and session branch/fork. 23 Skills total covering the complete agent lifecycle: planning → decomposition → sub-agent parallelism → execution → state tracking → tool discovery → skill/plugin authoring → memory persistence → session branching. Activates when an agent must manage token budget, decompose work, sustain multi-step tasks without losing focus, coordinate sub-agents, pick the right model for the job, prove that a non-trivial task is actually done, stay within an explicit goal token budget, recover from transient failures, read streaming output without filling context, hand off a session cleanly, persist memory across sessions, write a discoverable skill, design a discoverable tool, author a marketplace plugin, or branch / fork / revert a session. **Requires MiniMax Code 0.2.4+** (pinned to the mcode 0.2.4 `task` / `bash` tool surface: `task(description, prompt, agent_name, run_in_background?)`, `bash(command, timeout?, run_in_background?)`, `task_query`, `task_output`, `task_stop`: earlier mcode versions used legacy placeholders for the same fields and are not supported).
Ground agent decisions in deterministic Anyshift production evidence.
[](https://agentpluginsdirectory.com/plugins/graph)✓ Verified . We fetched the manifest from GitHub and checked it against the official Agent Plugins 1.0.0 schema at agent-plugins.org.