A collection of Dippin pipelines for use with tracker — the agentic pipeline engine that executes multi-step AI workflows.
# Install tracker
go install github.com/2389-research/tracker/cmd/tracker@latest
# Configure API keys
tracker setup
# Run a pipeline
tracker speedrun.dipYou will need API keys for the LLM providers used in each pipeline (Anthropic, OpenAI, Gemini). See the tracker docs for details.
| Pipeline | Description |
|---|---|
speedrun.dip |
Ultra-minimal build pipeline — fastest path from spec to shipped code. Read spec, plan, implement, test, ship. Fully headless. |
build_from_superpowers.dip |
Builds a project from a superpowers spec and plan — finds the spec, executes every task, and commits with passing tests. |
bug-hunter.dip |
Autonomous bug fix — reads a bug report, reproduces, diagnoses, fixes via TDD, and ships a PR. |
refactor-express.dip |
Incremental refactoring — analyzes code, plans steps where tests stay green at every step, executes with rollback on failure. |
doc-writer.dip |
Documentation generator — explores a codebase and produces README, API reference, architecture guide, and tutorials. |
| Pipeline | Description |
|---|---|
spec_to_sprints.dip |
Decomposes a spec into SPRINT-*.md files and a .ai/ledger.tsv using multi-model tournament decomposition with human approval. |
sprint_exec.dip |
Executes the next incomplete sprint from the ledger through implementation, validation, multi-model review, and completion. |
sprint_runner.dip |
Runs all sprints in sequence, looping until every sprint is completed. Inlines full sprint execution with review tournament and human gates. |
sprint_exec-cheap.dip |
Budget variant of sprint execution using smaller models (Haiku/Nano/Flash-Lite) with escalation. |
sprint_runner-cheap.dip |
Budget variant of the sprint runner with the same loop-and-escalation pattern. |
megaplan.dip |
Creates a sprint plan using multi-model orientation, drafting, critique, and merge stages. |
| Pipeline | Description |
|---|---|
spec_to_dip.dip |
Generates a validated .dip pipeline from a spec using multi-model tournament with domain-specific review panels. |
pipeline_from_spec.dip |
Generates a pipeline .dip file from a spec, scoring against objective pattern and coverage metrics. |
pipeline_from_spec_v2.dip |
Revised pipeline-from-spec with updated quality gates. |
pipeline_from_superpowers.dip |
Generates a pipeline from a superpowers-format spec. |
| Pipeline | Description |
|---|---|
20q.dip |
20 Questions game — the AI asks yes/no questions to guess what you're thinking of. |
story-engine.dip |
Choose-your-own-adventure — AI writes branching narrative scenes, you make choices that shape the plot. |
model-debate.dip |
Multi-model debate arena — Claude, GPT, and Gemini argue positions on a topic across rounds, then you judge. |
Each .dip file defines a workflow in the Dippin language — a DSL for describing agentic pipelines. Workflows declare nodes (agents, tools, human gates, parallel branches, conditionals) and edges with optional conditions.
Tracker reads the .dip file, builds the execution graph, and orchestrates LLM agents through it — dispatching tasks to Claude, GPT, or Gemini in isolated git worktrees with parallel execution support and a TUI dashboard.
- agent — LLM call with tool access
- human — Human input gate (choice, freeform, or interview)
- tool — Shell command execution
- parallel / fan_in — Parallel fan-out and synchronization
- conditional — Branching based on context
- tracker — The runtime engine that executes these pipelines
- dippin-lang — The Dippin language compiler, LSP, and toolchain
Built by 2389.ai