protodog
v3.7.0Protodog, the Claude-native agentic execution protocol: Task/Program entry skills, ideation audit dispatch, working-canvas and plan maintenance skills, the pinned challenge agent, enforcement hooks, validators, templates, and the guarded external-audit launcher.
By Pablo BelinLicense: Proprietary0 GitHub starsUpdated last month
Directory evidence
- Runtimes
- Claude Code
- Parsed components
- 6 skill or MCP entries
- Source updated
- Aug 22, 2026
- Manifest status
- Canonical path parsed
The directory validates manifest shape and source location. It does not execute the plugin or provide a security endorsement. Review the indexing methodology →
Install protodog for Claude Code
claude plugin marketplace add IchenDEV/agent-plugin-mkt
claude plugin marketplace update agent-plugin-marketplace
claude plugin install protodog@agent-plugin-marketplacePaste and run these commands in a terminal with Claude Code. They add and refresh the PluginsMP catalog, then install this plugin.
The installer fetches third-party code from the source repository shown on this page. This directory validates manifest structure and source location, but does not perform a security audit; review the manifest, components, and source before installing.
Get the source manually
git clone https://github.com/pablofrommars/protodogClone the source repository, then follow its setup instructions to add the plugin to a compatible client. The plugin root is protodog/.
Plugin files
├── .claude-plugin/plugin.json├── skills/deferred-review/SKILL.md├── skills/ideation-audit/SKILL.md├── skills/plan-sweep/SKILL.md├── skills/program/SKILL.md├── skills/task/SKILL.md└── skills/working-canvas/SKILL.md
Included Skills6
Review plan/deferred.md against current repository state — find rows closed by later work, fired triggers, and claims that no longer hold — and apply only engineer-ratified transitions. Use when the register may have drifted, including when a completed profile offered this review.
Dispatch independent adversarial audits of the current ideation session's settled conclusions — grounded against the repository, purely conceptual, or both — so design concepts and logic are settled before planning. Use from inside an ideation mini protocol.
Engineer-triggered retirement sweep of the plan/ tree — proves terminal, lifted, and committed with the deterministic checker, reviews citation liveness, and deletes only engineer-confirmed directories. Use after the commit to main, normally on main.
Start or resume the Protodog Program profile for a long-lived multi-track objective. Use when the engineer invokes the Program entry with intent, input/spec references, or an existing program plan.
Start or resume the Protodog Task profile for a bounded engineering objective. Use when the engineer invokes the Task entry with intent, input/spec references, or an existing task plan.
Maintain the working canvas — a session-scoped visual model of diagram-suited work (a mapping between structures, a topology, a lifecycle, or a data flow) — from the baseline turn onward, through ideation, planning, and execution. Use when such a session begins, when a decision surface or gate needs its options drawn in context, or when progress recoloring is due.
Plugin manifests1
{
"name": "protodog",
"displayName": "Protodog",
"description": "Protodog, the Claude-native agentic execution protocol: Task/Program entry skills, ideation audit dispatch, working-canvas and plan maintenance skills, the pinned challenge agent, enforcement hooks, validators, templates, and the guarded external-audit launcher.",
"version": "3.7.0",
"author": {
"name": "Pablo Belin"
},
"license": "Proprietary"
}For maintainers
If you maintain this plugin, link to this source-backed listing from your README so users can review its manifest and indexed components.
[protodog on Agent Plugins Marketplace](https://pluginsmp.com/plugins/protodog)