docs: add Fireworks AI integration page#846
Conversation
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
WalkthroughThis PR adds documentation for the Fireworks AI integration in Notte. It registers a new Estimated code review effort🎯 1 (Trivial) | ⏱️ ~5 minutes Possibly related PRs
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
This comment has been minimized.
This comment has been minimized.
|
| Filename | Overview |
|---|---|
| docs/src/integrations/fireworks.mdx | New integration page for Fireworks AI; ships with two known issues: missing model IDs for GLM-5 and MiniMax M2.5, and multiple placeholder URLs that may not yet be live. |
| docs/src/docs.json | Adds integrations/fireworks entry to the Integrations nav group between openai-cua and stagehand; straightforward and correct. |
Prompt To Fix All With AI
Fix the following 2 code review issues. Work through them one at a time, proposing concise fixes.
---
### Issue 1 of 2
docs/src/integrations/fireworks.mdx:55-59
**Missing model IDs for GLM-5 and MiniMax M2.5**
The table lists these two models but omits their `reasoning_model` strings. A user who wants to reproduce results with GLM-5 or MiniMax M2.5 has no actionable value to paste into the SDK call — they'd need to hunt the Fireworks catalogue themselves. The PR description acknowledges these IDs are unverified, but publishing the page without them means the docs are immediately incomplete for two of the three advertised models. Either add the verified IDs (e.g. `fireworks_ai/accounts/fireworks/models/<id>`) or hold the models until the IDs are confirmed.
### Issue 2 of 2
docs/src/integrations/fireworks.mdx:14-15
**Unresolved placeholder URLs**
Both `https://fireworks.ai/blog/agents-fail-on-execution` (used three times: lines 14, 66, 79) and `https://notte.cc/blog/agent-execution-tax` (line 15, 80) are cited as not-yet-live in the PR description. If either URL 404s at publication time, all outbound links from this page break and the benchmark claims become unverifiable. These should be confirmed live before the page is published, or swapped for temporary anchors.
Reviews (1): Last reviewed commit: "docs: add Fireworks AI integration page" | Re-trigger Greptile
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/src/integrations/fireworks.mdx`:
- Around line 56-63: Update the models table rows for GLM-5 and MiniMax M2.5 to
include the full LiteLLM/Notte reasoning_model strings like the Kimi row;
specifically change the GLM-5 row to include
`fireworks_ai/accounts/fireworks/models/glm-5` and the MiniMax M2.5 row to
include `fireworks_ai/accounts/fireworks/models/minimax-m2p5` so the
validated-models table uses canonical model IDs and avoids copy/paste failures.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: a376f204-7e0b-4f26-b3fa-82648712429d
📒 Files selected for processing (2)
docs/src/docs.jsondocs/src/integrations/fireworks.mdx
|
Preview deployment for your docs. Learn more about Mintlify Previews.
💡 Tip: Enable Workflows to automatically generate PRs for you. |
2027 — Eval complete for commit
|
| URL | Mapping |
|---|---|
| docs.notte.cc | → notte-docs-fireworks-integration.mintlify.app |
2027 // Getting Started — C 65.8/100
- █████████████░░░░░░░ ▼ -6.3 pts vs baseline| Time | Cost | Errors | Interruptions |
|---|---|---|---|
| 2m 34s ▲ +10s | $2.01 ▼ -$0.01 | 1 ▲ +1 | 1 |
Friction points:
- 🔴 SDK README and/or docs document
timeoutMinutesas a session option, but the API rejects it withextra_forbidden; the correct field isidle_timeout_minutes— . This mismatch directly caused a runtime error. - 🔴 No Node — .js / JavaScript code example showing how to get the CDP URL from a session and connect Playwright to it. Only Python examples exist in the quickstart, leaving JS developers to reverse-engineer from TypeScript types.
- 🔴 The canonical docs URL (docs — .notte.cc) issues a 308 permanent redirect to a Mintlify staging subdomain (notte-docs-fireworks-integration.mintlify.app), which looks unintentional and undermines trust.
- 🟡 notte — .cc/llms.txt exists but its content appears to redirect readers to llms-full.txt; a proper concise llms.txt summary distinct from the full context file would better serve LLM pre-context use cases.
- 🟡 Notte is not indexed on Context7, reducing discoverability for AI coding agents that rely on Context7 for up-to-date library documentation — .
Prompt
Use Notte to run a headless browser session. Follow the quickstart at https://notte.cc/ to set up a project that:
- Creates a cloud browser session
- Connects to it using Playwright, Puppeteer, or the Notte SDK
- Navigates to https://news.ycombinator.com
- Extracts the titles of the top 5 stories
- Prints them to stdout
Commit 6c63609 · View report → · View trace →
Evaluating agent experience using 2027.dev · View dashboard
2027 — Waiting for deployment on commit
|
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/src/integrations/fireworks.mdx`:
- Line 10: The docs incorrectly recommend a reasoning_model string with an
unsupported provider; update the Fireworks docs text so the example uses a
provider that matches the SDK's LlmProvider enum (or remove the explicit
provider prefix) instead of "fireworks_ai/accounts/fireworks/models/<id>" so
validation in packages/notte-core/src/notte_core/common/config.py (which derives
provider via model.split("/")[0]) will pass; reference the LlmProvider enum and
the provider parsing logic in config.py when changing the example or add a note
that this format is only valid once LlmProvider gains "fireworks" or
"fireworks_ai".
In `@docs/src/testers/integrations/fireworks_basic.py`:
- Line 9: The reasoning_model value uses an invalid provider ("fireworks_ai")
causing LlmModel.get_provider() (triggered when client.Agent() is instantiated)
to raise ValueError; update the reasoning_model string in fireworks_basic.py to
use a supported provider/model format (one of the providers in the LlmProvider
enum used by the SDK) or swap to a known-good model identifier (e.g., a provider
present in the SDK) so provider extraction succeeds and client.Agent() no longer
fails validation.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: fdf81240-ce1f-4481-8b27-63b465fa16f6
📒 Files selected for processing (3)
docs/src/integrations/fireworks.mdxdocs/src/snippets/integrations/fireworks_basic.mdxdocs/src/testers/integrations/fireworks_basic.py
✅ Files skipped from review due to trivial changes (1)
- docs/src/snippets/integrations/fireworks_basic.mdx
| with client.Session() as session: | ||
| agent = client.Agent( | ||
| session=session, | ||
| reasoning_model="fireworks_ai/accounts/fireworks/models/kimi-k2p5", |
There was a problem hiding this comment.
should be moonshotai/kimi-k2.6
There was a problem hiding this comment.
auctually I just reverted to use local notte instead of notte-sdk
Adds the Fireworks AI integration page to the docs and registers it under the Integrations group, following the existing openai-cua.mdx format. Companion page for the joint Notte x Fireworks benchmark launched last week. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Swap placeholder benchmark URL to the live Fireworks slug (agent-execution-tax) verified against fireworks.ai/blog. - Add canonical Fireworks model IDs for GLM-5 (glm-5) and MiniMax M2.5 (minimax-m2p5), verified against the Fireworks model catalogue at fireworks.ai/models. Resolves Greptile and CodeRabbit feedback on missing IDs and unresolved placeholder URLs. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The Fireworks integration page has inline Python examples for the same reason kernel.mdx does (integration-specific code that isn't worth extracting into a tester snippet). Add it to the same exception set so test_no_snippets_outside_folder passes. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Resources section now points to console.notte.cc (the product surface) instead of the source repo, and adds Browser Arena as an open benchmark reference. Both URLs verified live. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Move Python example to docs/src/testers/integrations/fireworks_basic.py and generate the corresponding snippet via make sniptest. The snippet is now imported as <FireworksBasic /> and runs through the standard tester test suite (DOCS_TEST_FAST_MODE syntax-checks it in CI). - Remove fireworks.mdx from the inline-snippet exception list now that the snippet pattern is in place. - Drop the "Validated Fireworks models" section (benchmark-specific numbers) and the "Reproducing the benchmark" pytest block. Keep the benchmark as a single link in the Overview, per CTO feedback. - Replace the table with a compact "Supported models" bullet list that preserves the actionable model IDs (Kimi K2.5, GLM-5, MiniMax M2.5) without the benchmark framing. Model string fireworks_ai/accounts/fireworks/models/kimi-k2p5 verified against the engine prefix in packages/notte-llm/src/notte_llm/engine.py and against the Fireworks model catalogue at fireworks.ai/models. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Trim marketing-style adjectives ("production-grade inference
infrastructure"), redundant preambles ("Set the following environment
variables:"), and "That's it" closing lines. Tighten section titles and
the Supported models description. Net effect: same information, ~30%
less prose, more conversational tone aligned with the stagehand and
massive integration pages.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
giordano-lucas
force-pushed
the
docs/fireworks-integration
branch
from
55f3cee to
aa2ec8c
Compare
|
|
Found 1 test failure on Blacksmith runners: Failure
|
2 participants
Summary
docs/src/integrations/fireworks.mdxdocumenting how to run Notte agents on Fireworks-served models (Kimi K2.5, GLM-5, MiniMax M2.5)docs/src/docs.jsonunder the Integrations group (placed afteropenai-cua)openai-cua.mdxformat: frontmatter +AgentMdNoticepartial +<Steps>setup + validated-models table + reproducibility block + resourcesWhy now
Companion docs page for the Fireworks ecosystem listing ask. Their agent-frameworks page links out to each framework's own Fireworks integration docs (LangChain, LlamaIndex, and Pydantic all follow this pattern). Notte needs its own equivalent page live before pinging them to add Notte to the listing, otherwise the card has nothing to link to.
Verified for merge
kimi-k2p5,glm-5,minimax-m2p5), verified against the Fireworks model catalogue.https://fireworks.ai/blog/agent-execution-tax.https://notte.cc/blog/agent-execution-taxis live (published 2026-05-19).Test plan
🤖 Generated with Claude Code
Summary by CodeRabbit