You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
vscode-extension: v0.4.5 — new mascot + configurable spend cap + i2i timeout
UX
- New mascot artwork on the empty state: a bigger AI+coin themed
Franklin pixel-art portrait. The PNG ships with a flood-filled
alpha channel so the mascot composites directly onto whichever
theme background sits behind the panel — no rounded-rectangle
frame, no mix-blend-mode hack. Original PNG (with the dark frame)
also kept on disk as franklin-mascot.png for record / fallback.
- Settings popover gains a "Per-turn spend cap (USD)" field. Empty
keeps the existing $0.25 default; "0" disables the cap entirely;
any positive number sets it. Local mirror of core PR #20.
Core (cherry-picks of in-flight PRs)
- src/tools/imagegen.ts: image-to-image gets a 180s timeout (was
shared 60s with text-to-image — gpt-image-2 edits on a few-MB
reference image consistently hit AbortError before completion).
Mirror of PR #19.
- src/agent/loop.ts + src/commands/config.ts: MAX_TURN_SPEND_USD
is no longer hard-coded; reads max-turn-spend-usd from config
with fallback to $0.25. Mirror of PR #20.
Bumps vscode-extension to 0.4.5; updates README changelog.
Copy file name to clipboardExpand all lines: vscode-extension/README.md
+6Lines changed: 6 additions & 0 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -41,6 +41,12 @@ Franklin is an autonomous AI agent that runs directly in VS Code. It doesn't jus
41
41
42
42
## Changelog
43
43
44
+
### 0.4.5
45
+
-**New mascot artwork on the empty state** — bigger, transparent-background AI+coin themed Franklin pixel-art portrait (no more dark rounded-rectangle frame, blends seamlessly into any theme background)
46
+
-**Per-turn spend cap is now configurable** — new ⚙️ settings field "Per-turn spend cap (USD)" lets you raise the default \$0.25 limit (or set 0 to disable) without editing source; mirrors the new `franklin config set max-turn-spend-usd <n>` CLI key
47
+
-**Image-to-image timeout fix** — `gpt-image-2` reference-image edits no longer abort after 60s (the old shared budget couldn't cover reasoning-driven edits + base64 upload + x402 retry); image-to-image now gets 180s, text-to-image keeps 60s
48
+
- Synced with Franklin core: PR #19 (i2i timeout), PR #20 (configurable spend cap), PR #21 (README VS Code section)
49
+
44
50
### 0.4.3
45
51
-**History replay shows generated media inline** — closing and reopening a conversation now re-renders any images / videos as preview cards instead of dropping them
46
52
-**"+" New chat truly resets the session** — previously only the UI cleared while the agent kept the same `sessionId`, leaking tool guards (`ImageGen disabled`) and prior context into what looked like a new chat
@@ -180417,7 +180418,10 @@ Open with: open ${outPath}${contentSummary}`
180417
180418
} catch (err) {
180418
180419
const msg = err.message || "";
180419
180420
if (msg.includes("abort")) {
180420
-
return { output: "Image generation timed out (60s limit). Try a simpler prompt.", isError: true };
180421
+
return {
180422
+
output: referenceImage ? "Image-to-image timed out (180s limit). The reference image may be too large or the model under load \u2014 try a smaller/simpler image." : "Image generation timed out (60s limit). Try a simpler prompt.",
0 commit comments