Never stop coding. Free MIT AI gateway: one endpoint, 375 providers (150+ free), 1,400+ models Kimi, Claude, GPT, Gemini, GLM, DeepSeek, MiniMax. Works with Claude Code, Codex, Cursor, OpenCode, Cline & Copilot. Quota-aware auto-fallback, RTK+Caveman compression saves 15-95% tokens, MCP/A2A, Desktop/PWA. Built by 900+ contributors
TypeScript a2aai-agentsai-gatewayanthropicclaudeclaude-codeclinecodexcopilotcursordeepseekfree-aigeminikimillm-gatewaymcpopenaiopenai-proxyqwentoken-saver
chore: bump version to 3.7.9-pre
Antigravity Assistant85f2f8d8d8f9
diff --git a/package-lock.json b/package-lock.jsonindex 12fafde14c..59b1726b69 100644--- a/package-lock.json+++ b/package-lock.json@@ -1,12 +1,12 @@{"name": "omniroute",- "version": "3.7.7",+ "version": "3.7.8","lockfileVersion": 3,"requires": true,"packages": {"": {"name": "omniroute",- "version": "3.7.7",+ "version": "3.7.8","hasInstallScript": true,"license": "MIT","workspaces": [diff --git a/package.json b/package.jsonindex 18e9704d0d..41cec7a351 100644--- a/package.json+++ b/package.json@@ -1,6 +1,6 @@{"name": "omniroute",- "version": "3.7.7",+ "version": "3.7.8","description": "Smart AI Router with auto fallback — route to FREE & cheap models, zero downtime. Works with Cursor, Cline, Claude Desktop, Codex, and any OpenAI-compatible tool.","type": "module","bin": {diff --git a/scripts/scratch/categorize_ideas.py b/scripts/scratch/categorize_ideas.pynew file mode 100644index 0000000000..943939d630--- /dev/null+++ b/scripts/scratch/categorize_ideas.py@@ -0,0 +1,39 @@+import os+import shutil+import subprocess++categories = {+ "viable": [1718, 1731, 1764],+ "need_details": [1765, 1679, 1594, 1584],+ "defer": [1845, 1833, 1814, 1786, 1737, 1736, 1735, 1716, 1591, 1590, 1589, 1588, 1587],+ "notfit": [1826, 1788, 1529, 1586] # 1586 is already exists, goes to notfit+}++# Create dirs+os.makedirs("_ideia/viable/need_details", exist_ok=True)+os.makedirs("_ideia/defer", exist_ok=True)+os.makedirs("_ideia/notfit", exist_ok=True)++files = os.listdir("_ideia")++for f in files:+ if not f.endswith(".md"): continue+ num_str = f.split('-')[0]+ if not num_str.isdigit(): continue+ num = int(num_str)++ src = os.path.join("_ideia", f)++ if num in categories["viable"]:+ dst = os.path.join("_ideia/viable", f)+ elif num in categories["need_details"]:+ dst = os.path.join("_ideia/viable/need_details", f)+ elif num in categories["defer"]:+ dst = os.path.join("_ideia/defer", f)+ elif num in categories["notfit"]:+ dst = os.path.join("_ideia/notfit", f)+ else:+ continue++ shutil.move(src, dst)+ print(f"Moved {f} to {dst}")diff --git a/scripts/scratch/harvest_issues.py b/scripts/scratch/harvest_issues.pynew file mode 100644index 0000000000..649e75f511--- /dev/null+++ b/scripts/scratch/harvest_issues.py@@ -0,0 +1,96 @@+import json+import os+import subprocess+import re++issues = [1845, 1833, 1826, 1814, 1788, 1786, 1765, 1764, 1737, 1736, 1735, 1731, 1718, 1716, 1679, 1594, 1591, 1590, 1589, 1588, 1587, 1586, 1584, 1529]++os.makedirs('_ideia', exist_ok=True)++def slugify(value):+ value = re.sub(r'\[Feature\]', '', value, flags=re.IGNORECASE)+ value = re.sub(r'[^\w\s-]', '', value).strip().lower()+ return re.sub(r'[-\s]+', '-', value)++for num in issues:+ print(f"Processing #{num}...")+ try:+ res = subprocess.run(+ ['gh', 'issue', 'view', str(num), '--repo', 'diegosouzapw/OmniRoute', '--json', 'number,title,labels,body,comments,createdAt,author,assignees'],+ capture_output=True, text=True, check=True+ )+ data = json.loads(res.stdout)++ slug = slugify(data['title'])+ filename = f"_ideia/{data['number']}-{slug}.md"++ author = data['author']['login']+ created_at = data['createdAt']+ body = data['body']+ comments = data['comments']++ participants = set([author])+ comments_str = ""+ if comments:+ for c in comments:+ c_author = c['author']['login']+ participants.add(c_author)+ comments_str += f"**@{c_author}** ({c['createdAt']}):\n{c['body']}\n---\n"+ else:+ comments_str = "*No comments.*"++ participants_list = "\n".join([f"- @{p}" for p in participants])++ content = f"""# Feature: {data['title']}++> GitHub Issue: #{data['number']} — opened by @{author} on {created_at}+> Status: 📋 Cataloged | Priority: TBD++## 📝 Original Request++{body}++## 💬 Community Discussion++{comments_str}++### Participants++{participants_list}++### Key Points++- Needs detailed analysis++## 🎯 Refined Feature Description++Feature needs manual refinement and interpretation to fill logical gaps and outline high-level technical scope.++### What it solves++- TBD++### How it should work (high level)++1. TBD+2. TBD++### Affected areas++- TBD++## 📎 Attachments & References++- Check issue body for references++## 🔗 Related Ideas++- None yet+"""+ with open(filename, 'w', encoding='utf-8') as f:+ f.write(content)++ except Exception as e:+ print(f"Error processing {num}: {e}")++print("Done.")diff --git a/scripts/scratch/refine_viable.py b/scripts/scratch/refine_viable.pynew file mode 100644index 0000000000..abdf86ed5d--- /dev/null+++ b/scripts/scratch/refine_viable.py@@ -0,0 +1,91 @@+import os++viable_data = {+ 1718: {+ "title": "expose upstream error details in client-facing error responses",+ "solves": "Debugging upstream errors is difficult because they are hidden behind generic '[400] Error from provider' wrappers.",+ "how": "1. Modify `buildErrorBody` to accept `upstream_details`.\n2. In `BaseExecutor` or specific handlers, parse the raw upstream response on failure.\n3. Propagate the parsed body to the client response inside `upstream_details`.",+ "areas": "open-sse/executors/base.ts, open-sse/utils/errors.ts, src/app/api/v1/chat/completions/route.ts",+ "req_summary": "Expose the upstream error body (e.g. context_length_exceeded) directly in the error response under an `upstream_details` key, without breaking OpenAI compatibility.",+ "req_approach": "Modify the central error generation functions (like `buildErrorBody`) to optionally accept an `upstreamDetails` object. Update the request executors to pass the JSON parsed error from the upstream response into this new parameter when a request fails.",+ "req_files": "| File | Changes |\n|---|---|\n| `open-sse/utils/errors.ts` | Update `buildErrorBody` to include `upstream_details`. |\n| `open-sse/executors/base.ts` | Extract response body on failure and pass to error builder. |",+ "req_effort": "Low. A few files changed. No breaking changes."+ },+ 1731: {+ "title": "(combo): provider-level exhaustion tracking to skip same-provider targets",+ "solves": "Combo routing wastes significant time retrying multiple targets from the same provider when the entire provider is rate-limited or quota-exhausted.",+ "how": "1. Track 429 quota exhaustion errors at the provider level.\n2. In `combo.ts`, before attempting a target, check if its provider is currently marked as exhausted.\n3. If exhausted, skip the target and move to the next provider.",+ "areas": "open-sse/services/combo.ts, open-sse/services/accountFallback.ts",+ "req_summary": "Implement provider-level 429 exhaustion tracking in the combo router so it skips remaining targets of a provider if a 429 quota exhaustion occurs.",+ "req_approach": "Add a temporary exclusion set in `handleComboChat` that tracks providers that have returned a hard 429. Before evaluating the next target in the combo, check if its provider is in the exclusion set and skip it if true.",+ "req_files": "| File | Changes |\n|---|---|\n| `open-sse/services/combo.ts` | Add logic to track provider failures and skip matching targets. |\n| `open-sse/services/accountFallback.ts` | Properly bubble up the 429 status. |",+ "req_effort": "Medium. Needs careful state tracking across the combo loop. No breaking changes."+ },+ 1764: {+ "title": "Make installation script detect termux",+ "solves": "OmniRoute fails to start or install properly on Termux because `wreq-js` attempts to load a native `libgcc` arm64 module which is incompatible.",+ "how": "1. Update the `postinstall` script to check `process.env.PREFIX` for termux.\n2. If termux is detected, gracefully skip or patch the wreq-js installation/loading.",+ "areas": "scripts/postinstall.mjs, open-sse/executors/wreq.ts",+ "req_summary": "Detect termux environments during installation or runtime and gracefully handle the `wreq-js` native module failure, allowing the rest of OmniRoute to function.",+ "req_approach": "Modify `scripts/postinstall.mjs` or the wreq-js loader logic. If `process.env.PREFIX && process.env.PREFIX.includes('termux')` is true, avoid hard crashing on wreq-js load failures.",+ "req_files": "| File | Changes |\n|---|---|\n| `scripts/postinstall.mjs` | Add termux detection and warning. |\n| `open-sse/utils/env.ts` (or similar) | Graceful downgrade. |",+ "req_effort": "Low. Very localized fix."+ }+}++for root, _, files in os.walk("_ideia/viable"):+ for f in files:+ if not f.endswith(".md") or "requirements" in f: continue+ num = int(f.split('-')[0])+ if num in viable_data:+ path = os.path.join(root, f)+ with open(path, 'r') as file:+ content = file.read()++ d = viable_data[num]++ # replace TBDs+ content = content.replace("### What it solves\n\n- TBD", f"### What it solves\n\n- {d['solves']}")+ content = content.replace("### How it should work (high level)\n\n1. TBD\n2. TBD", f"### How it should work (high level)\n\n{d['how']}")+ content = content.replace("### Affected areas\n\n- TBD", f"### Affected areas\n\n- {d['areas']}")+ content = content.replace("Feature needs manual refinement and interpretation to fill logical gaps and outline high-level technical scope.", "Refined and scoped for implementation.")++ with open(path, 'w') as file:+ file.write(content)++ req_content = f"""# Requirements: {d['title']}++> Feature Idea: [#{num}](./{f})+> Research Date: 2026-05-01+> Verdict: ✅ VIABLE++## 🔍 Research Summary++{d['req_summary']}++## 📚 Reference Implementations++| # | Repository | Stars | Last Updated | Approach | Relevance |+| --- | ---------------- | ----- | ------------ | -------- | ------------ |+| 1 | OmniRoute Source | - | 2026-05-01 | Internal | High |++## 📐 Proposed Solution Architecture++### Approach++{d['req_approach']}++### Modified Files++{d['req_files']}++## ⚙️ Implementation Effort++- **Estimated complexity**: {d['req_effort']}+- **Breaking changes**: No+"""+ req_path = os.path.join(root, f.replace(".md", ".requirements.md"))+ with open(req_path, 'w') as req_file:+ req_file.write(req_content)++ print(f"Refined {num} and created requirements.")