Current tier, loaded profiles, credential and optional-key state, source health, database info, and what the next tier would add. With verify=true each stored credential is tested against its live API.

Checks Keychain access, the SQLite database and any credentials an existing Heimdall install may already hold (never used without approval). Lists the manual setup links.

Lists every profile, whether it is loaded, and the tools it contains.

Searches the whole tool catalog (including tools that are not loaded or not implemented yet) by name, description or profile.

Lists the data sources with their tier, required credentials, health status and confidence level. With probe=true it makes one live request per source instead of reporting what past calls recorded.

Explains how a score is computed: formula, inputs and caveats.

The full App Report in one call: listing + metadata audit, review mining, keyword opportunity (when terms are given), market difficulty, ASC performance (from previously imported data), rank history, seasonal calendar and culture notes. Sections degrade independently — a broken source is reported as skipped, never invented.

Deep-dive on ONE competitor: public listing profile, changes observed between captures (inferred, low confidence), monetization signals (price + SensorTower estimates), keyword portfolio against your own ranks, and complaint mining with verbatim quotes. Everything is public data — what cannot be observed is skipped, not guessed.

Side-by-side comparison of your app against the SELECTED competitors on the shortlist, across metadata, product page, monetization and keyword ranks. The feature axis is semantic and is derived by you from the descriptions — the data carries the raw text.