Tools provided
Tools the agent exposes — your AI client calls them automatically when it needs them.
- run_full — Orchestrated path: from the provided repo scan and target locales, extract translation keys, build the locale files, check coverage, flag right-to-left risk, list missing translations and plan the PR.
- analyze — Run pure deterministic analysis over provided scan data with no model call and return scored findings for coverage, right-to-left risk and the marketing-copy guardrail.
- extract_keys — Extract one translation-key entry per hardcoded string — the key, source text, source file, whether it is auto-translatable and any applied translations.
- build_locale_files — Build the i18next-compatible locale JSON files, one per requested locale.
- check_coverage — Return the per-locale translation-coverage findings only — how many UI strings are translated for each requested locale.
- check_rtl_risk — Return the right-to-left layout-risk findings only — strings long enough to overflow once the layout mirrors, which need bidi verification.
- list_missing_translations — Return the missing-translation report per locale, separating dictionary gaps from marketing copy deliberately withheld for a human translator.
- plan_pr — Return the translation-PR plan: branch name, PR title and the list of files changed.
- list_capabilities — List the agent's static capabilities: the available tools and the analyzer's coverage and right-to-left thresholds. Useful for discovery.
- plan_inputs — Plan the inputs for a tool: returns the clarifying questions, the JSON schema and a ready-to-edit example.
- discover_intent — Understand your goal and co-design the exact input through clarifying questions before running.
- open_form — Opens the guided form to collect the repo scan, the target locales and the scope. The entry point for running the agent on your own data.
- run_form — Internal: invoked by the guided form when the user submits. Not for direct use.
