aindex autopilot "fix billing login bug" --json
Autopilot · For: agent
AI coding sessions waste tokens because the first move is usually broad grep, repeated reads, and whole-file context.
Autopilot turns a task into an ordered context plan: budget, pack, files, risks, owners, tests and do-not-read zones.
$ aindex autopilot "fix billing login bug" --json
naive_tokens
required_pack_tokens
read_first
tests_to_runCommand
Agent Cost Autopilot
Give the agent a budgeted plan before it starts reading.
context action=autopilot
naive_tokensrequired_pack_tokensread_firsttests_to_run
What the Brain does automatically
The Brain fires this workflow for you.
The first response tells the team how much context the agent would have burned and what it should read instead.
implement
When you ask an agent to build a feature.
fix
When you ask an agent to fix a bug.
start
At the start of every agent session.
public/assets/benchmarks.jsonRequired tier
Where this becomes commercially available.
Turns a 180k-token naive read into a 9.4k budgeted pack before the agent spends a single token.
Beta
Usable for pilots, with contracts still being hardened from observed runs.
Pilot-grade
Buyer-safe for pilots, but still below fully hardened Sourcegraph-class parity.
- Setup success
- CLI autopilot
- MCP context
- ROI dashboard
- Pilot report
This workflow stays discoverable across setup, CLI, MCP status, docs, dashboard and pricing — never a hidden command.
Feature visibility checklist
Buyer-visible names covered by Agent Cost Autopilot.
This keeps grouped product names discoverable without claiming a separate maturity level from the underlying workflow.
Agent Cost AutopilotBeta
Agent Cost Autopilot
Budgeted plan before the agent starts reading.
- Surface
- CLI, MCP context, dashboard, pilot report
- Buyer-safe status
- Beta · Pilot-grade
- Runtime
- preview only
Autopilot
Start every agent run with a cost-aware execution plan.
10k tokens instead of 200k