@clickminded/coauthor-cli 0.0.88 → 0.0.90

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -22,6 +22,8 @@ export declare function draftCompletionMessage(base: string, status: JsonValue):
22
22
  * question. Nothing when the plan has no AI search data.
23
23
  */
24
24
  export declare function seoPlanAiSearchLines(value: unknown): string[];
25
+ /** "AI search mentions since 2026-11-18: Google AI answers 482 (up 12), ChatGPT 37 (down 3)." Null without a comparison. */
26
+ export declare function seoPlanAiTrendLine(value: unknown): string | null;
25
27
  /**
26
28
  * `--cta save,follow,custom:Join the webinar` -> the web app's cta_variants
27
29
  * list (at most 4 endings, at most one custom).
package/dist/commands.js CHANGED
@@ -2708,11 +2708,32 @@ function seoPlanResultsMessage(data) {
2708
2708
  if (missing.length > RESULT_ROWS_SHOWN)
2709
2709
  lines.push(` and ${missing.length - RESULT_ROWS_SHOWN} more (see --json data.missing_urls).`);
2710
2710
  }
2711
+ const aiTrend = seoPlanAiTrendLine(lastCheck?.ai_visibility);
2712
+ if (aiTrend)
2713
+ lines.push('', aiTrend);
2711
2714
  const reason = seoPlanCheckReason(lastCheck);
2712
2715
  if (reason)
2713
2716
  lines.push('', reason);
2714
2717
  return lines.join('\n');
2715
2718
  }
2719
+ /** "AI search mentions since 2026-11-18: Google AI answers 482 (up 12), ChatGPT 37 (down 3)." Null without a comparison. */
2720
+ export function seoPlanAiTrendLine(value) {
2721
+ if (!value || typeof value !== 'object')
2722
+ return null;
2723
+ const trend = planRecord(value);
2724
+ const since = planString(trend, 'previous_at');
2725
+ if (!since)
2726
+ return null;
2727
+ const parts = [];
2728
+ for (const [key, label] of [['google', 'Google AI answers'], ['chat_gpt', 'ChatGPT']]) {
2729
+ const row = trend[key] && typeof trend[key] === 'object' ? planRecord(trend[key]) : null;
2730
+ if (!row || typeof row.mentions !== 'number' || typeof row.change !== 'number')
2731
+ continue;
2732
+ const change = row.change === 0 ? 'no change' : `${row.change > 0 ? 'up' : 'down'} ${Math.abs(row.change)}`;
2733
+ parts.push(`${label} ${row.mentions} (${change})`);
2734
+ }
2735
+ return parts.length > 0 ? `AI search mentions since ${since.slice(0, 10)}: ${parts.join(', ')}.` : null;
2736
+ }
2716
2737
  function seoPlanWaitFailure(status) {
2717
2738
  const code = planString(status, 'error_code');
2718
2739
  const message = planString(status, 'error_message') ?? 'The plan run failed.';
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@clickminded/coauthor-cli",
3
- "version": "0.0.88",
3
+ "version": "0.0.90",
4
4
  "type": "module",
5
5
  "bin": {
6
6
  "coauthor": "dist/index.js",
@@ -90,7 +90,7 @@ Some plans tag ideas whose searches are ones where Google AI answers cite a comp
90
90
  - `ai_search_volume`: DataForSEO's estimate of monthly AI searches behind the idea's matching searches.
91
91
  - `questions`: on United States / English plans, up to 3 ChatGPT questions about the idea whose answers cite a competitor (`question`, `competitor`). Empty elsewhere.
92
92
 
93
- An AI gap idea went through the same checks as every other idea; the tag explains why it matters, it does not mean the idea is more likely to rank. The human output of `show` prefixes it with `[AI gap]` and names the competitors. When you write one, `seo-plan write` puts its questions into the post's brief, and the outline answers them.
93
+ An AI gap idea went through the same checks as every other idea; the tag explains why it matters, it does not mean the idea is more likely to rank. The human output of `show` prefixes it with `[AI gap]` and names the competitors. When you write one, `seo-plan write` puts its questions into the post's brief. The outline checks that its sections cover what those answers need, without adding a FAQ or question headings.
94
94
 
95
95
  `ai_search_update_available` (top level of `show --json`) is `true` for a full plan made before AI search ideas went live on 2026-11-18. That plan has no `ai_visibility` and no AI gap ideas. Running `coauthor seo-plan generate` updates it and adds them, using the user's credits (`estimate_credits`). Mention it once when it is `true`, and do not update the plan without the user's go-ahead. It turns `false` after an update, and it is always `false` for previews and stale plans.
96
96
 
@@ -128,7 +128,8 @@ Posts written without a target search (mostly drafted before SEO mode) are track
128
128
 
129
129
  - `data.results` has one row per tracked post (a live address and a search), best news first (moved up, then page one by position, then the rest, then posts no check has read yet): `project_id`, `title`, `keyword`, `tracked_url`, `position`, `previous_position`, `movement`, `ranking_url`, `checked_at` and `tracking`. `tracking` is `null` for a post written for a target search; for a post written without one it is `{keyword, source}`: the search tracked now (`keyword` is `""` while the next check will pick one; it can differ from the row's `keyword`, which is what the last reading used) and who chose it (`source`: `ranked` or `suggested` = CoAuthor picked it, `manual` = the user did, `null` = not picked yet). `position` is Google's organic position; page one means 10 or better. It is `null` when the post's address is not within the top results the check looked at (`data.last_check.depth`, 30 by default), and `movement` is then `not_found`; if `ranking_url` is set, another page of the same site ranks for that search instead, which is worth telling the user. `movement` is `up` or `down` only for a move of 3 places or more, or a move onto or off page one; `same` is steady; `new` means the first reading of that search and address, so there is nothing to compare with yet; `pending` means the post went live after the last check and has no reading yet (`position` and `checked_at` are `null`; the next check reads it).
130
130
  - `data.missing_urls` lists posts that have no live address (`project_id`, `title`, `keyword`, `needs_search`, `publication_status`: `needs_search` is `true`, with `keyword` `""`, for a post written without a target search that has no search to track yet, so the next check picks one; `publication_status`: `published`, `submitted`, or `null` when nobody marked the post); only posts explicitly marked as still being edited are left out. Most posts are unmarked, so a `null` post may not be live yet: ask the user before setting an address. They are not tracked until you set one with `coauthor post-details set <project-id> --live-url <url>` (add `--track-search "<search>"` to choose the search for a post without a target search); the next check picks it up.
131
- - `data.last_check` is `{status, finished_at, partial, partial_reason, plan_refresh_skipped, error_code, credits_charged, free, depth}`. `partial` means the check stopped early (for example credits ran out); `plan_refresh_skipped` (`stale` or `failed`) means the plan's own chances were not refreshed because the plan needs updating, so run `coauthor seo-plan generate`. `status` `skipped` with `error_code` `insufficient_credits` means the account had too few credits and nothing was spent.
131
+ - `data.last_check` is `{status, finished_at, partial, partial_reason, plan_refresh_skipped, error_code, credits_charged, free, depth, ai_visibility}`. `partial` means the check stopped early (for example credits ran out); `plan_refresh_skipped` (`stale` or `failed`) means the plan's own chances were not refreshed because the plan needs updating, so run `coauthor seo-plan generate`. `status` `skipped` with `error_code` `insufficient_credits` means the account had too few credits and nothing was spent.
132
+ - `data.last_check.ai_visibility` is how many AI answers mentioned the site at that check, next to the previous reading: `{fetched_at, previous_at, google, chat_gpt}`. Each platform is `{mentions, cited, previous_mentions, change}` or `null` (`chat_gpt` is `null` outside United States / English). `previous_at` is `null` on the first reading, and `change` is then `null`. The previous reading is the last check's, or the plan's when the plan was updated since. The whole field is `null` when the check did not read AI mentions. Report the change as it is. A change of a few mentions is normal month to month, so don't explain it as the effect of a post. The human output of `results` prints one line for it.
132
133
  - Everything is empty or `null` while monthly checks are not switched on for the account, and before the first check has run.
133
134
 
134
135
  When you report results, lead with what moved up and what reached page one, then name any posts without a live address. Do not call a `same` post a failure: one monthly reading is noisy.
@@ -695,7 +695,7 @@ Some plans tag ideas whose searches are ones where Google AI answers cite a comp
695
695
  - `ai_search_volume`: DataForSEO's estimate of monthly AI searches behind the idea's matching searches.
696
696
  - `questions`: on United States / English plans, up to 3 ChatGPT questions about the idea whose answers cite a competitor (`question`, `competitor`). Empty elsewhere.
697
697
 
698
- An AI gap idea went through the same checks as every other idea; the tag explains why it matters, it does not mean the idea is more likely to rank. The human output of `show` prefixes it with `[AI gap]` and names the competitors. When you write one, `seo-plan write` puts its questions into the post's brief, and the outline answers them.
698
+ An AI gap idea went through the same checks as every other idea; the tag explains why it matters, it does not mean the idea is more likely to rank. The human output of `show` prefixes it with `[AI gap]` and names the competitors. When you write one, `seo-plan write` puts its questions into the post's brief. The outline checks that its sections cover what those answers need, without adding a FAQ or question headings.
699
699
 
700
700
  `ai_search_update_available` (top level of `show --json`) is `true` for a full plan made before AI search ideas went live on 2026-11-18. That plan has no `ai_visibility` and no AI gap ideas. Running `coauthor seo-plan generate` updates it and adds them, using the user's credits (`estimate_credits`). Mention it once when it is `true`, and do not update the plan without the user's go-ahead. It turns `false` after an update, and it is always `false` for previews and stale plans.
701
701
 
@@ -733,7 +733,8 @@ Posts written without a target search (mostly drafted before SEO mode) are track
733
733
 
734
734
  - `data.results` has one row per tracked post (a live address and a search), best news first (moved up, then page one by position, then the rest, then posts no check has read yet): `project_id`, `title`, `keyword`, `tracked_url`, `position`, `previous_position`, `movement`, `ranking_url`, `checked_at` and `tracking`. `tracking` is `null` for a post written for a target search; for a post written without one it is `{keyword, source}`: the search tracked now (`keyword` is `""` while the next check will pick one; it can differ from the row's `keyword`, which is what the last reading used) and who chose it (`source`: `ranked` or `suggested` = CoAuthor picked it, `manual` = the user did, `null` = not picked yet). `position` is Google's organic position; page one means 10 or better. It is `null` when the post's address is not within the top results the check looked at (`data.last_check.depth`, 30 by default), and `movement` is then `not_found`; if `ranking_url` is set, another page of the same site ranks for that search instead, which is worth telling the user. `movement` is `up` or `down` only for a move of 3 places or more, or a move onto or off page one; `same` is steady; `new` means the first reading of that search and address, so there is nothing to compare with yet; `pending` means the post went live after the last check and has no reading yet (`position` and `checked_at` are `null`; the next check reads it).
735
735
  - `data.missing_urls` lists posts that have no live address (`project_id`, `title`, `keyword`, `needs_search`, `publication_status`: `needs_search` is `true`, with `keyword` `""`, for a post written without a target search that has no search to track yet, so the next check picks one; `publication_status`: `published`, `submitted`, or `null` when nobody marked the post); only posts explicitly marked as still being edited are left out. Most posts are unmarked, so a `null` post may not be live yet: ask the user before setting an address. They are not tracked until you set one with `coauthor post-details set <project-id> --live-url <url>` (add `--track-search "<search>"` to choose the search for a post without a target search); the next check picks it up.
736
- - `data.last_check` is `{status, finished_at, partial, partial_reason, plan_refresh_skipped, error_code, credits_charged, free, depth}`. `partial` means the check stopped early (for example credits ran out); `plan_refresh_skipped` (`stale` or `failed`) means the plan's own chances were not refreshed because the plan needs updating, so run `coauthor seo-plan generate`. `status` `skipped` with `error_code` `insufficient_credits` means the account had too few credits and nothing was spent.
736
+ - `data.last_check` is `{status, finished_at, partial, partial_reason, plan_refresh_skipped, error_code, credits_charged, free, depth, ai_visibility}`. `partial` means the check stopped early (for example credits ran out); `plan_refresh_skipped` (`stale` or `failed`) means the plan's own chances were not refreshed because the plan needs updating, so run `coauthor seo-plan generate`. `status` `skipped` with `error_code` `insufficient_credits` means the account had too few credits and nothing was spent.
737
+ - `data.last_check.ai_visibility` is how many AI answers mentioned the site at that check, next to the previous reading: `{fetched_at, previous_at, google, chat_gpt}`. Each platform is `{mentions, cited, previous_mentions, change}` or `null` (`chat_gpt` is `null` outside United States / English). `previous_at` is `null` on the first reading, and `change` is then `null`. The previous reading is the last check's, or the plan's when the plan was updated since. The whole field is `null` when the check did not read AI mentions. Report the change as it is. A change of a few mentions is normal month to month, so don't explain it as the effect of a post. The human output of `results` prints one line for it.
737
738
  - Everything is empty or `null` while monthly checks are not switched on for the account, and before the first check has run.
738
739
 
739
740
  When you report results, lead with what moved up and what reached page one, then name any posts without a live address. Do not call a `same` post a failure: one monthly reading is noisy.