If you already use Claude Code, Cline, Codex, or OpenCode
The management layer above your coding agent.
Keep using the coding agent you know. aidd adds dependency-ordered backlogs, bounded multi-iteration runs, and a local panel for reviewing progress across projects. Acceptance criteria and run history stay with each repo.
When one project becomes several
Your CLI sessions can have their own instructions, plans, and history. Managing several projects adds another set of questions: which feature is ready to run, what is blocked, and which results still need review?
aidd keeps that work in a shared backlog and run view. You choose the agent for each run and can reuse the same workflow across repositories.
What aidd is
aidd is a local orchestrator - a Bun/TypeScript CLI plus a local web panel -
that launches your agent as the execution engine. A spec becomes features with
acceptance criteria and dependencies under .aidd/features/. A run
loops: pick the next eligible feature, compile the prompt, invoke the CLI once,
check the result. Repeat until the backlog empties or the run hits its iteration
or wall-clock limit.
Plan
Spec to dependency-ordered features, each with acceptance criteria.
Build
Iterations, one CLI invocation each, watchable live from the panel.
Verify
The agent runs the project's required tests, lint, typecheck, or browser checks and reports the results.
Record
Run ledger, per-iteration logs, and audit reports as plain files.
aidd checks completion claims against feature status and flags reports that admit skipped live verification. It does not independently rerun every check: you still need to review the results and their coverage.
What you get on top
Your CLI stays
Use Claude Code, Cline, Codex, OpenCode, KiloCode, Grok, an OpenAI-compatible API, or local Ollama / LM Studio. The backend is a per-run setting.
Records in the repo
.aidd/runs.jsonl, per-iteration logs, feature status, and
audit reports live next to the code as plain files. You can search and
diff them without aidd; the format is documented.
Audits with evidence
Built-in audits cover security, architecture, performance, and integration. Findings must cite evidence from the current code before they are recorded as backlog work.
A director for the fleet
Discovers every project under the roots you configure, scans them in cycles, and proposes prioritized work. Accept a suggestion and it launches a run. Cycles can run on a schedule, and suggestion auto-launch is available: bounded, and off by default.
Repeatable workflows
Multi-step workflows - review, fix, verify - saved as JSON files and rerun on any project, from the panel or on a schedule.
MCP and Telegram
An MCP server (bun run start -- --mcp, alongside the
running panel) exposes aidd to other agent sessions. The Telegram bridge
lets you talk to the Director from your phone after you connect your own
bot and allow-list your chat. Both are included.
What it isn't
You can still use your CLI directly for interactive or exploratory work. aidd is
a local application with no hosted control plane. It is opinionated about
project state: work is tracked as features in .aidd/, and coding
runs select from that backlog rather than freeform prompts - a one-off
instruction is its own mode (--prompt), not the default. Windows 11
with PowerShell 7 is the first-class platform; Linux runs the full quality gate
in CI, and macOS is expected to work but is checked by hand. There are no
prebuilt binaries - aidd ships as source on every platform and runs on Bun.
Local runs, explicit limits. Runs use your existing subscription or API keys, with iteration and wall-clock limits. Cost and token ceilings surface as warnings in the run log; they do not stop a run. aidd stores its records locally, has no account or hosted control plane, and sends no telemetry. Coding runs send prompts and relevant project content to your chosen backend. Configured integrations and downloads you initiate can also make network requests.