Editorial Policy

CarryFeed pages distinguish product capability, source observation, and editorial recommendation. A product capability must exist in the working product. A source observation must stay within the public example or documented sample. An editorial recommendation is presented as a choice to test, not a fact about every account.

• Product capability: what CarryFeed can do now, including the input, result, failure state, and user confirmation required.
• Source observation: what a named public link, account history sample, product state, or primary reference actually shows.
• Editorial recommendation: an interpretation or next step that may be useful, with uncertainty and account context left intact.

Viewer, search, statistics, trend, and downloader pages are checked against the public source types they describe. We record the input used, the result that appears, and the limits that change the user's next step.

• Start with a public x.com or twitter.com URL, public handle, search, or supported public result.
• Describe only the profile, post, article, media, search, trend, or sampled-history context that the source exposes.
• Label examples as examples; do not present fabricated UI or generated content as a live result.
• Retest affected pages when product behavior or the public platform changes.

CarryFeed's job is to help shorten the path from a useful signal to a thoughtful post and a consistent publishing practice. Growth still depends on the account, audience, point of view, execution, and factors outside CarryFeed's control.

• We do not promise followers, reach, engagement, leads, virality, or a particular publishing outcome.
• Timing guidance is based on the public-history sample available for the account; it is evidence for a decision, not a prediction of future performance.
• Content, audience-fit, and workflow recommendations are hypotheses to review and test, not private platform intelligence.
• Private-beta capabilities and future product direction are labeled separately. Roadmap ideas are not described as available features.

CarryFeed can use the context a user provides, saved writing preferences, and public source material to propose posts, threads, and rewrites. The proposal remains visible as an AI-assisted revision until the user decides what to keep.

• Voice Memory records explicit preferences and accepted direction; it does not establish authorship or permission to impersonate someone else.
• AI output must not invent quotes, experience, research, customer results, or product evidence.
• Content and media checks flag observable issues such as length, links, image requirements, alt text, or style signals. They do not certify truth, originality, legal compliance, or likely performance.
• The user remains responsible for factual review, rights, disclosure, tone, and the final point of view.

Connected-account actions are different from public-link research. Before CarryFeed schedules or publishes, the user sees the content and exact action and gives explicit approval.

• A prepared time or draft does not schedule or publish itself.
• Failed and reconciliation-required deliveries remain visible instead of being reported as successful.
• CarryFeed does not claim autonomous replies, follows, likes, direct messages, or hands-off audience growth.
• A successful delivery confirms the requested platform action; it is not an endorsement by X or a promise of distribution.

Public research and connected-account workflows have different permissions, but neither is permission to bypass platform controls.

• Private, protected, deleted, suspended, restricted, or login-only sources are not represented as public results.
• CarryFeed does not bypass account privacy settings, paywalls, technical restrictions, or platform access controls.
• Public search, recommendations, trends, and account-history samples are not described as complete coverage of X.
• Media availability, result detail, and browser save behavior can vary by source, device, browser, and platform behavior.

We review affected product and editorial copy when CarryFeed capabilities or material platform behavior changes. If a claim is no longer supported, the useful response is to narrow or remove it before expanding the promise.