Upgrade to Goose v1.46.0 and use per-message cost UI with cache-token tracking
Ship Goose v1.46.0 so sessions show per-message tokens/cost/TTFT/tok/s, cache-token-aware totals, streaming shell output, and cache-safe append-only request assembly from the unrolled agent loop.Why this loop
v1.46.0 ships the unrolled agent loop (#9574) and cache-safe request assembly: append-only turn context with declared cache semantics (#11022). Cost reporting now tracks cache tokens (#9752), records per-message usage/cost with derived session totals (#10172), and the UI shows tokens, cost, TTFT, and tok/s on each message (#10210). Shell output streams while commands run (#10808). Upgrade so live sessions use this request path and the per-message cost UI. bodyOmitted is 40; do not assume those unlisted changes.
Proposed actions
- Upgrade Goose to v1.46.0 from https://github.com/aaif-goose/goose/releases/tag/v1.46.0 and stop if
goose --versiondoes not report v1.46.0. - In a v1.46.0 session, run at least three billed model turns and capture evidence that each message shows tokens, cost, TTFT, and tok/s (#10210), session totals are derived from those per-message records (#10172), and cache tokens are included in cost reporting (#9752).
- Run a 5+ turn task on a cache-capable provider. Do not rebuild the full prompt each turn. Keep turn context append-only with declared cache semantics (#11022) on the unrolled agent loop (#9574). Afterward, report per-message cache-token vs non-cache cost so totals match cache-aware billing.
- From the same session, run a shell command that prints for at least 10 seconds and fail the check if output appears only after exit—stdout must stream while the command is still running (#10808).
Agent prompt
Paste into your agent or query via MCP (get_agent_prompt) — free, no extra AI costGoose session task
DevAgentRadar → Goose
You are helping me adopt a real coding-assistant change. Work only from the facts below. Do not invent features.
Context
Assistant: Goose Proposal: Upgrade to Goose v1.46.0 and use per-message cost UI with cache-token tracking Summary: Ship Goose v1.46.0 so sessions show per-message tokens/cost/TTFT/tok/s, cache-token-aware totals, streaming shell output, and cache-safe append-only request assembly from the unrolled agent loop. Primary source: https://github.com/aaif-goose/goose/releases/tag/v1.46.0
Why it matters
v1.46.0 ships the unrolled agent loop (#9574) and cache-safe request assembly: append-only turn context with declared cache semantics (#11022). Cost reporting now tracks cache tokens (#9752), records per-message usage/cost with derived session totals (#10172), and the UI shows tokens, cost, TTFT, and tok/s on each message (#10210). Shell output streams while commands run (#10808). Upgrade so live sessions use this request path and the per-message cost UI. bodyOmitted is 40; do not assume those unlisted changes.
Suggested actions
-
Upgrade Goose to v1.46.0 from https://github.com/aaif-goose/goose/releases/tag/v1.46.0 and stop if
goose --versiondoes not report v1.46.0. -
In a v1.46.0 session, run at least three billed model turns and capture evidence that each message shows tokens, cost, TTFT, and tok/s (#10210), session totals are derived from those per-message records (#10172), and cache tokens are included in cost reporting (#9752).
-
Run a 5+ turn task on a cache-capable provider. Do not rebuild the full prompt each turn. Keep turn context append-only with declared cache semantics (#11022) on the unrolled agent loop (#9574). Afterward, report per-message cache-token vs non-cache cost so totals match cache-aware billing.
-
From the same session, run a shell command that prints for at least 10 seconds and fail the check if output appears only after exit—stdout must stream while the command is still running (#10808).
Config surfaces this release may change
- background and headless runs — check your repo before applying
- slash commands — check your repo before applying
- context and compaction settings — check your repo before applying
- agent context files — check your repo before applying
- hooks — check your repo before applying
- MCP servers — check your repo before applying
- skills — check your repo before applying
After you finish
Do not report this as applied to DevAgentRadar. You cannot write the visitor's loop.
Tell the human: open https://devagentradar.com/proposals/goose-v1-46-0-upgrade-to-goose-v1-46-0-and-use-per-message-cost-ui-with- and mark Applied, Skipped, or Failed. Proposal id: 96ae9283-fa8b-46c6-a6ee-e5e730168440
Your job
- Restate the change in one sentence.
- Propose a minimal plan for my repo (or a throwaway pilot).
- Implement only what I approve; prefer small diffs and tests.
- Call out risks (permissions, breaking APIs, cost).
Start by confirming you understood the proposal.
Your loop
This browser · no sign-in · not shared as “you”After you run the prompt
Only you can mark this. Agents cannot write your loop.
Your decision stays on this device. A public tally appears after a few votes.