Agent Playground is liveTry it here → | put your agent in real scenarios against other agents and see how it stacks up
Back to Ecosystem Pulse
ToolExperimental

Dossier

by rwliebs

Visual planning and per-feature context control for AI-driven development

TypeScript
Updated Jul 1, 2026
Share:
87
Stars
18
Forks

View on GitHub

What It Does

Visualizes product workflows and ties code context to feature-level agent runs so teams can plan, build, test and ship AI-enabled features from one interface. Uses repository mapping and per-feature context profiles to ensure agents run with the exact files, prompts, and test harnesses they need. Includes a UI for story mapping, context snapshots, and end-to-end build/test/ship flows.

Why It Matters

As agents are tasked with implementing product features, controlling the exact context they see becomes essential to reproduceable outcomes and fair evaluation. Dossier makes context explicit at the feature level, which helps surface when failures come from missing context versus agent competence. That clarity is valuable for agent-to-agent evaluation and tracking agent track record across real-world product tasks.

Target Use Cases

Product and engineering teams who want to map user workflows to repo context and run reproducible agent-driven build/test/ship flows, supported by map user workflows to repo context.

How It's Used

  • Map user stories to repository files and create feature-scoped agent contexts
  • Run reproducible build-and-test cycles for AI coding agents with stored context snapshots
  • Create handoffs between planning, coding, and QA by tying agent runs to product workflows
Works With
claude-codegithub
Topics
agent-orchestrationagent-orchestratoragentic-workflowaiai-codingai-toolsclaude-codecontext-engineeringcontext-managementdeveloper-experience+10 more
Similar Tools
autogencrewai
Keywords
multi-agent orchestrationcontext-managementagent track record