AiderDesk

You will like this if you have used Aider and wished the session had a window. The trait that decides it in daily use is that every task gets its own git worktree, so a run you dislike is a directory you delete rather than a mess in your checkout, and you can leave two tasks going without them stepping on each other. Projects and tasks give the day a shape that a terminal history does not.

Pick it if you want to supervise more than one thing at a time. Pick Aider when the terminal was never the problem, or Cline if you would rather stay inside VS Code.

The risk is surface area against maintainer count. Thirty-plus providers, two MCP directions, profiles, subagents and an API is a matrix no small team tests exhaustively, and the row confirms containers are documented for running the app rather than for confining the agent. Shell commands therefore execute in your own environment. Isolation protects your files and not your machine, and those are different guarantees.

Done right: nothing lands unseen. Proposed changes are gated behind a diff review and a tool-approval prompt, so the dangerous default is opt-in rather than assumed.

Two design notes. 1. The context engine pairs embeddings with a repository map, which hedges the known weakness of each: retrieval misses structure, structure misses intent. That is a considered choice rather than an accidental one. 2. Edit application is inherited from the upstream project this began as, so its behaviour under model change is already characterised elsewhere.

No numbers are offered here, and none are borrowed either, which is more restraint than most vendors manage. The documentation describes capability and declines to quantify it, so a reader should treat the claims as documented intent, not demonstrated performance.

There is no revenue line here and no attempt at one: permissive licence, no paid tier, no hosted edition. For a consultancy that is rational, since a widely installed tool is a pipeline of clients who already run your code. Pricing power is nil, but nothing is being priced.

The dependency worth watching is nominal. This product borrows the recognition of the command-line tool it wraps, so its demand curve is not fully its own. Likely outcome is continued sponsorship while it serves the consultancy, and a slow stop when it does not. Position: adopt, expect no roadmap commitments, and do not pay for support that is not sold.

Licence cost across sixty engineers is zero and the model spend lands on whichever provider keys we issue, which at least makes the meter ours to watch. The REST API is the part that matters to me, since it means an integration is possible without a human clicking. Against that: the row records no single sign-on, no SCIM and no audit trail, and self-hosting the browser UI puts a service on my platform team's plate with no vendor to call.

Onboarding is half a day for anyone who has run the underlying tool. Approved with conditions: one team hosts it, and provider keys are issued centrally.

Apache-2.0 with no contributor agreement theatre, installed by brew, scoop or npm depending on what mood I am in. The provider list runs from Bedrock and Azure through OpenRouter and LiteLLM and finishes at Ollama and LM Studio, which is the entry that matters: the model can live on my own hardware and the app talks to localhost. MCP works in both directions, so it consumes my servers and exposes itself to other clients.

Forking is realistic because it is TypeScript and Electron, not a compiled black box. This is a tool I could keep alive alone.