Copy
Copies only the displayed code or formula. AI explanations are excluded.
Choose a workflow or click a control in the simulated extension. The tour changes the mock screen, animates the next action, and explains the result.
Copies only the displayed code or formula. AI explanations are excluded.
Combines the answer with a stated requirement while preserving and validating its language.
Boosts the result in search and puts it in the Favorites filter.
Adds curated answers locally. Personal snippets support editing and deletion.
Sends guarded context to Ollama. Pasted content is marked as untrusted data.
Runs selectors only against inert pasted HTML before generating code.
Press Ctrl+Shift+K and describe the desired outcome. Exact titles, aliases, language shortcuts, tags, favorites, notes, personal snippets, and imported code-pack content influence ranking. Arrow keys select results and Enter opens one.
Select webpage text and right-click. Explain Selection creates a guarded explanation, Explain Error starts diagnosis, and Save Snippet opens a prefilled form. DevRecall listens even while the panel is already open and never continuously reads the page.
Save titles, languages, folders, tags, aliases, notes, and code. Export snippets as JSON or Markdown. Optional code packs contain validated references and learning cards and remain absent until imported.
Manual mode provides Previous, Skip, and Next controls for a chosen language or pack. Auto mode uses Chrome alarms to mark cards due from every 15 minutes to weekly without unexpectedly opening the panel.
Code, HTML, errors, and retrieved text are wrapped as untrusted data. Wrong-language output is retried once and rejected if still invalid. Extractor scripts are scanned for network, storage, navigation, execution, and DOM-mutation operations.
DevRecall defaults to http://localhost:11434. Start Ollama, then open Settings and select Discover models.
winget install Ollama.Ollama
ollama serve
ollama pull qwen2.5-coder:7b
[Environment]::SetEnvironmentVariable(
"OLLAMA_ORIGINS",
"chrome-extension://*",
"User"
)
# Restart Ollama.
brew install ollama
launchctl setenv OLLAMA_ORIGINS "chrome-extension://*"
ollama serve
ollama pull qwen2.5-coder:7b
curl -fsSL https://ollama.com/install.sh | sh
systemctl --user edit ollama
# Add under [Service]:
Environment="OLLAMA_ORIGINS=chrome-extension://*"
systemctl --user daemon-reload
systemctl --user restart ollama
ollama pull qwen2.5-coder:7b
chrome-extension://YOUR_EXTENSION_ID where supported.ollama signin
ollama run gpt-oss:120b-cloud
Cloud models require confirmation for every request because data leaves the device and charges or account limits may apply.