tdlp
v0.5.0Evidence-grounded engineering for an LLM agent: consistency-belief (does the design follow?), component-belief (does the built thing work?), stamp-monitor (is the evidence still current?), their two workflow skills, and the file-blind consistency verifier.
By David Chen0 GitHub starsUpdated 2 hours ago
Directory evidence
- Runtimes
- Claude Code
- Parsed components
- 6 skill or MCP entries
- Source updated
- Sep 30, 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 tdlp for Claude Code
claude plugin marketplace add IchenDEV/agent-plugin-mkt
claude plugin marketplace update agent-plugin-marketplace
claude plugin install tdlp@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/davechendatascience/Theoretically_Driven_LLM_PlanningClone the source repository, then follow its setup instructions to add the plugin to a compatible client. The plugin root is plugin/.
Plugin files
├── .claude-plugin/plugin.json├── skills/adopt-goals/SKILL.md├── skills/component-belief/SKILL.md├── skills/consistency-belief/SKILL.md└── .mcp.json
Included Skills3
Set up goals.yaml in a project that already uses TDLP -- draft the human's goals, the interfaces between them and how each is measured, for the human to review and commit. Use when the user asks to set up, add, draft or adopt goals, or when status(view="goals") says no goals are in effect.
Ground engineering claims in measured evidence via the component-belief MCP. Use before reporting that something works, when diagnosing which component is the bottleneck, when deciding whether a change is ready, or whenever you are about to state a number you did not measure.
Check that a design claim is logically sound before (or while) building it, via the consistency-belief MCP -- axioms, definitions, lemmas and branches in consistency.yaml, verified by falsification probes served by status(view="probe"). Use when proposing a design or rule change, when a design claim is about to be stated as true, when an axiom or lemma is edited, or when asked whether the logic holds.
MCP servers3
- command
- uvx
- args
- --from git+https://github.com/davechendatascience/[email protected] component-belief-mcp
- env.BELIEF_PROJECT_ROOT
- ${CLAUDE_PROJECT_DIR}
- env.BELIEF_ACTOR
- agent
- env.UV_LINK_MODE
- copy
- command
- uvx
- args
- --from git+https://github.com/davechendatascience/[email protected] consistency-belief-mcp
- env.CONSISTENCY_PROJECT_ROOT
- ${CLAUDE_PROJECT_DIR}
- env.CONSISTENCY_ACTOR
- agent
- env.UV_LINK_MODE
- copy
- command
- uvx
- args
- --from git+https://github.com/davechendatascience/[email protected] stamp-monitor-mcp
- env.STAMP_MONITOR_ROOT
- ${CLAUDE_PROJECT_DIR}
- env.UV_LINK_MODE
- copy
Plugin manifests1
{
"name": "tdlp",
"version": "0.5.0",
"description": "Evidence-grounded engineering for an LLM agent: consistency-belief (does the design follow?), component-belief (does the built thing work?), stamp-monitor (is the evidence still current?), their two workflow skills, and the file-blind consistency verifier.",
"author": {
"name": "David Chen"
},
"repository": "https://github.com/davechendatascience/Theoretically_Driven_LLM_Planning"
}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.
[tdlp on Agent Plugins Marketplace](https://pluginsmp.com/plugins/tdlp)