opencode-see-image 0.8.5 → 0.9.0

package/index.ts

@@ -217,6 +217,7 @@ function readProviderKey(providerID: string): string | null {
 
 async function seeImageViaSDK(
   client: any,
+  $: any,
   dataUrl: string,
   mediaType: string,
   prompt: string,
@@ -231,7 +232,27 @@ async function seeImageViaSDK(
     fs.writeFileSync(tmpPath, Buffer.from(b64, "base64"))
   } catch {}
 
-  const fileUrl = `file://${tmpPath}`
+  // For free opencode models, use CLI instead of SDK (SDK returns empty)
+  const freeFallback = async (modelID: string, userPrompt: string): Promise<string | null> => {
+    try {
+      const controller = new AbortController()
+      const timer = setTimeout(() => controller.abort(), TIMEOUT)
+      const proc = $`opencode run -f ${tmpPath} -m opencode/${modelID} ${userPrompt} --format json --dangerously-skip-permissions`
+      const out = await proc.text()
+      clearTimeout(timer)
+      for (const line of out.split("\n").filter(Boolean)) {
+        try {
+          const parsed = JSON.parse(line)
+          if (parsed?.part?.type === "text" && parsed?.part?.text) {
+            return parsed.part.text
+          }
+        } catch {}
+      }
+    } catch {}
+    return null
+  }
+
+  const fileUrl = tmpPath
   let result: { text: string; model: string; provider: string } | undefined
 
   try {
@@ -245,6 +266,17 @@ async function seeImageViaSDK(
     candidates.push({ providerID: "opencode", modelID: "mimo-v2.5-free" })
 
     for (const { providerID, modelID } of candidates) {
+      if (providerID === "opencode") {
+        // SDK session.prompt returns empty for free models; use CLI instead
+        const text = await freeFallback(modelID, prompt)
+        if (text) {
+          result = { text, model: modelID, provider: providerID }
+          break
+        }
+        errors.push(`${providerID}/${modelID}: no text from CLI fallback`)
+        continue
+      }
+
       let sessionID: string | undefined
       try {
         const sessionRes = await client.session.create({ body: {} })
@@ -454,6 +486,7 @@ const SeeImagePlugin: Plugin = async (ctx) => {
       } else {
         result = await seeImageViaSDK(
           client,
+          $,
           resolved.dataUrl,
           resolved.mediaType,
           prompt,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "opencode-see-image",
-  "version": "0.8.5",
+  "version": "0.9.0",
   "description": "Give non-vision opencode models the ability to see images/screenshots by routing them to a vision-capable model (MiniMax M3 via opencode-go by default).",
   "type": "module",
   "main": "index.ts",