Route and author durable long-horizon loop-graph runs. Use for multi-round work needing durable state, gated milestones, owner boundaries, host switching, or independent audit. Route code cleanup to /loop-converge, feature requirements to /loop-deliver, and evidence-backed solution comparison to /loop-research; use loop-graph directly only for a custom run shape. For a self-contained task, recommend the host's ordinary task or goal directly. Do not execute or resume generated runtime node files.
Scanned 8/30/2026
Install to Claude Code
npx -y skills add levi-qiao/longgraph-skill --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of longgraph?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/levi-qiao-longgraph)More formats (shields.io, HTML) on the badges page.
---
name: longgraph
description: Route and author durable long-horizon loop-graph runs. Use for multi-round work needing durable state, gated milestones, owner boundaries, host switching, or independent audit. Route code cleanup to /loop-converge, feature requirements to /loop-deliver, and evidence-backed solution comparison to /loop-research; use loop-graph directly only for a custom run shape. For a self-contained task, recommend the host's ordinary task or goal directly. Do not execute or resume generated runtime node files.
---
# longgraph — route to the right durable run
longgraph has one shared compiler, [`loop-graph`](skills/loop-graph/SKILL.md),
and focused entries that bind only goal-specific rules. All compile the same
executor, supervisor, ledger, directives, and ops artifacts. Do not execute the
generated runtime nodes from this authoring skill.
## Fit and route
- Use [`loop-converge`](skills/loop-converge/SKILL.md) for multi-round unused
code, duplication, consolidation, reuse, or slimming.
- Use [`loop-deliver`](skills/loop-deliver/SKILL.md) for a product or engineering
requirement that needs incremental implementation and acceptance proof.
- Use [`loop-research`](skills/loop-research/SKILL.md) to compare feasible
approaches with open-source implementations, primary research, and controlled
experiments before committing to one.
- Use [`loop-graph`](skills/loop-graph/SKILL.md) directly only when none of the
focused packs fits but the work still needs durable state, gates, owner
boundaries, host switching, or independent verification.
- If the request is one self-contained task that fits a normal host goal/task,
explain that longgraph adds no value and recommend sending the task directly. Do
not wrap it in another objective or invent a lighter longgraph mode.
Shared reference: [`lib/methodology.md`](lib/methodology.md) explains why the graph's
discipline exists. The baseline and pack boundary is in
[`skills/loop-graph/docs/preset-contract.md`](skills/loop-graph/docs/preset-contract.md).
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!