Compare dated Akta and public-web signals across a reviewed competitor set.
Compare dated Akta and public-web signals across a reviewed competitor set.
--- name: akta-competitor-change-tracker description: Compare private-company competitors across reviewed Akta data and bounded public-web evidence without overstating incomplete or conflicting signals. --- # Competitor Change Tracker ## Loadout capabilities Use <action-tag>api:akta_api:company_search</action-tag> to resolve every competitor. Use <action-tag>api:akta_api:company_metrics_fixed</action-tag> for a consistent metric snapshot, <action-tag>api:akta_api:activity_signals_variable</action-tag> for selected activity fields, <action-tag>api:akta_api:reviews_variable</action-tag> for chosen review evidence, and the News operation from <integration-tag>api:akta_api</integration-tag> for a bounded news interval. Use the Exa Search Action from <integration-tag>composio:exa_tools</integration-tag> only to find public evidence for changes that matter to the comparison, then use <action-tag>api:tinyfish_api:fetch_urls</action-tag> to extract the selected official or primary-source pages. ## Outcome Produce a dated competitor-change report with comparable observations, meaningful changes, confidence, unknowns, and credit use. The report is a snapshot; this Skill does not create a persistent monitor or claim a trend from one observation. ## Required inputs - Competitor names or domains and the comparison question. - Baseline date or prior reviewed snapshot when available. - Selected metric, activity, review, and news fields. - Date window, materiality threshold, permitted public-source scope, and maximum approved Loadout credit spend. ## Workflow 1. Define the comparison table and materiality rules before retrieval. Keep the same requested fields across competitors. 2. Resolve every competitor to an official domain and Akta identity. Exclude unresolved companies rather than comparing possible namesakes. 3. Preflight variable activity and review requests for the complete company set, add fixed metrics and bounded news, and show the total quote before execution. 4. For material changes that need public corroboration, run bounded Exa searches and fetch only selected official or primary-source URLs with Tinyfish. Apply the same source-selection rule across competitors. 5. Normalize returned dates and units. When a reviewed prior snapshot exists, calculate changes against it; otherwise label the result as a current baseline. 6. Rank only changes that meet the supplied materiality rule. Separate Akta values, public observations, calculated differences, interpretation, conflicting sources, and missing evidence. ## Output contract - `Scope`: companies, identities, time window, fields, and materiality rule. - `Comparison`: one row per company with comparable dated observations. - `Material changes`: prior value, current value, difference, Akta or public evidence, and confidence. - `News and reviews`: attributed signals, not generalized sentiment claims. - `Unknowns`: unresolved identities, missing fields, and incomparable periods. - `Cost`: quoted and actual credits plus Actions called. ## Failure and recovery - Stop priced requests for any ambiguous company and continue only if the remaining comparison still answers the question. - If there is no prior snapshot, establish a baseline and do not describe values as changes. - If the quote exceeds budget, reduce fields or companies with user approval; do not silently sample. - If sources disagree, show each observation and its date instead of selecting the most favorable value.