Research prompts

You are working in a repository where I have configured the OpenAI Codex CLI to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.

You are working in a repository where I have configured Claude Code to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.5. The bridge is the part that usually breaks. Confirm which protocol Claude Code speaks to it, and whether keep-alive lines from the API reach the client during a long read. If the first token takes more than a minute to arrive, the bridge is buffering and the client will abort the stream.

You are working in a repository where I have configured DeepSeek Harness to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.

You are working in a repository where I have configured Pi to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.

You are working in a repository where I have configured Oh my Pi to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.

You are working in a repository where I have configured GitHub Copilot in VS Code to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.5. This client can discover models from the API. Confirm whether it listed cog-fast and cog-pro by itself, or whether you had to declare them.

You are working in a repository where I have configured Zed to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.5. Zed hides the request shape behind one setting. Confirm whether the API is receiving max_tokens or max_completion_tokens, and set capabilities.max_tokens_parameter accordingly.

You are working in a repository where I have configured Continue.dev to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.5. Confirm that agent mode works, which requires the tool_use capability to be declared explicitly. If tools are not offered to the model, describe what the client sent.

You are working in a repository where I have configured Cline to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.5. Use the extension's own Verify button first and tell me what it reported.

You are working in a repository where I have configured Roo Code to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.5. This extension uses native tool calling only, with no fallback. Ask it to use one tool and report whether the tool call arrived correctly formed. This is the check that matters for this client.

You are working in a repository where I have configured Aider to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.5. Confirm that the model string carries the openai/ prefix, and run one small edit task rather than only a chat request, because the edit path is where this client differs.

You are working in a repository where I have configured opencode to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.5. Confirm which AI SDK package the provider uses. It must be @ai-sdk/openai-compatible for chat completions; @ai-sdk/openai targets a different endpoint and will fail here.

You are working in a repository where I have configured Goose to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.5. This client reads OPENAI_HOST and OPENAI_BASE_PATH rather than the variable name other tools use. Confirm which one it actually read, and remember that it ignores keys placed in config.yaml.

You are working in a repository where I have configured Crush to use Cognitivers as its model provider.The API is OpenAI-compatible chat completions.- Base URL: https://api.cognitivers.com/v1- Streaming base URL, same API for very large inputs: https://stream.cognitivers.com/v1- Models: cog-fast (agents, pipelines, volume), cog-pro (one hard task), cog-embed (retrieval)- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these four things and then stop.1. Check that the credential is present in my environment and tell me the name of the variable you found it in. Do not print the key itself, not even partially.2. Send one non-streaming chat completion asking for a single short sentence, using cog-fast. Show me the exact command or code you used and the model's reply.3. Send the same request with streaming enabled and confirm you received server-sent events, not one buffered body.4. If any step fails, tell me the exact error, the status code and the request that produced it, then stop. Do not try more than two fixes, and do not change any configuration file without showing me the diff first.Rules: never write the key into a file that is tracked by git, never print it in full, and if the request works, say so in one line instead of writing a report.5. Confirm the provider type is openai-compat and not openai, and tell me whether the model discovery ran against the API or whether you registered the models by hand.

Cursor cannot use a third-party endpoint: its own documentation lists five providers that accept your own key and documents no base URL field. Do not spend time trying to make it work.Two things I do want from you, in the repository I am working in.1. Search the code for every place this project calls an OpenAI-compatible API (the OpenAI SDKs, the Vercel AI SDK, LangChain, LiteLLM, fetch against a /v1 path) and list the file and line of each one. I want to know what would have to change if a tool did support it.2. Check whether any of those call sites is Cursor-specific, for example an .cursor/rules file or a Cursor model name in a config, and tell me what it pins.Do not change any file. Report what you found, and if you find nothing, say that in one line.

You are working in a repository that calls the Cognitivers API, an OpenAI-compatible chat completions endpoint.- Base URL: https://api.cognitivers.com/v1- Models: cog-fast, cog-pro, cog-embed- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these five things and then stop.1. Find every place in this repository that constructs an API client for a hosted model, and tell me the file and the argument name each one uses for the base URL. I want to know whether they use base_url or baseURL, because one letter of case is the usual mistake.2. Confirm the credential is read from the environment, and name the variable. Do not print the key.3. Send one non-streaming request to cog-fast from the language this repository uses and show me the reply.4. Send the same request with streaming enabled and confirm you received server-sent events. If this repository uses the Vercel AI SDK, check whether the call goes through .chat(); the default factory targets the Responses API and would fail here.5. If any step fails, give me the exact error, the status code and the request that produced it, and stop. Do not change any file without showing me the diff first.If everything works, say so in one line.

You are working in a repository where a LiteLLM proxy is configured in front of Cognitivers.- Upstream base URL: https://api.cognitivers.com/v1- Model strings as LiteLLM needs them: openai/cog-fast, openai/cog-pro, openai/cog-embed- Authentication: an API key in the Authorization header as "Bearer sk-cog-..."Do these five things and then stop.1. Validate the config file and start the proxy. Show me the exact command.2. Send one chat request through the proxy to cog-fast and show the reply.3. Send one request with streaming enabled and confirm the events pass through the proxy without being buffered.4. If the proxy is also serving Claude Code, confirm which inbound protocol it accepts and send one request through that path too.5. If any step fails, give me the exact error, the status code and the request that produced it, and stop. Do not print the key in full and do not write it into a tracked file.If everything works, say so in one line.