@seoagent-official/seoagent 1.91.0 → 1.92.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@seoagent-official/seoagent",
-  "version": "1.91.0",
+  "version": "1.92.0",
   "description": "The persistent AI SEO agent for Claude Code. Audits, keyword strategy, briefs, articles, real product screenshots from your repo, and the autopilot loop (cloud detects → CLI executes → ack closes) — other SEO tools write the prompt, SEOAgent runs it.",
   "type": "module",
   "bin": {

package/skills/references/audit-checks.md

@@ -10,7 +10,7 @@ Loaded by Phase 1 (Technical SEO Audit) when running a full audit. The main `SKI
 
 **An incomplete capture makes every finding a LOWER BOUND.** If `evidence.md` contains a **"Pages NOT captured"** section (`capture_complete: false` in its frontmatter), the crawl discovered pages it could not fetch — after per-page retries and a sequential re-fetch. The rollup and per-page findings then describe only the captured subset: report the code-generated "N discovered pages could not be crawled — evidence is incomplete" finding verbatim, state the incompleteness (and each failure reason) in the audit and final summary, and never treat an uncaptured page as passing any check — no claim about it (present OR absent) is Confirmed.
 
-**The evidence covers the whole crawl, not just the homepage — audit accordingly.** The crawl discovers subpages from the live sitemap + homepage nav links (default 15 pages) and writes a per-page section for **EVERY page in `evidence.md`**, plus a **`## Site-wide rollup`** section that aggregates the per-page gaps: *Pages missing canonical*, *Pages missing meta description*, *Pages with multiple H1s*, *Pages with no structured data*, and *Images missing alt* (total + per-page counts with the offending srcs). Run the per-page checks below against **every crawled page** — a subpage's missing canonical or a blog post's missing meta description is exactly what the rollup lists — and report those findings from the rollup + per-page sections, never from homepage-only inspection or guesswork.
+**The evidence covers the whole crawl, not just the homepage — audit accordingly.** The crawl discovers subpages from the live sitemap + homepage nav links and fetches **every one of them** (safety cap 500; `--max` only ever narrows it) and writes a per-page section for **EVERY page in `evidence.md`**, plus a **`## Site-wide rollup`** section that aggregates the per-page gaps: *Pages missing canonical*, *Pages missing meta description*, *Pages with multiple H1s*, *Pages with no structured data*, and *Images missing alt* (total + per-page counts with the offending srcs). Run the per-page checks below against **every crawled page** — a subpage's missing canonical or a blog post's missing meta description is exactly what the rollup lists — and report those findings from the rollup + per-page sections, never from homepage-only inspection or guesswork.
 
 **The crawl also writes `.seoagent/audit/findings.md` — the complete, code-generated findings baseline. Your audit builds on it.** It derives one human-readable finding per confirmed rollup issue (affected URLs, why it matters, suggested fix, `Confirmed` label, `Evidence:` citation), including client-rendered shells, sitemap coverage gaps, robots.txt facts, and broken crawled URLs. Treat it as the completeness floor: `audit/latest.md` adds GSC cross-referencing, prioritization, and strategy ON TOP of findings.md — it must carry **every** finding forward (**never re-derive from scratch, never truncate**), and the final message lists every finding title (`seoagent summary` does this mechanically via its "Technical findings (from live crawl)" section). Session economy applies to bookkeeping, never to findings completeness. Items findings.md lists under **"Already present on the live site"** must never become "add X" recommendations — when the repo source lacks something the live page serves, the **repo source is stale — the live page already serves it; reconcile the source**.
 
@@ -317,7 +317,7 @@ Recommendation: "Add breadcrumb navigation + BreadcrumbList JSON-LD on deep page
 
 ### The crawl is the gate for ALL live-state work
 
-Any audit, technical-SEO review, "what's wrong with my site", or "add schema/meta/canonical" request starts with `seoagent crawl` — even a quick one, even when the repo source looks obvious. The crawl covers the homepage **plus subpages discovered from the live sitemap + homepage nav links** (default 15 pages, fetched concurrently), so it's fast and it sees the blog posts and subpages where the real gaps hide. **No live-state claim and no "add X" recommendation may be emitted unless `.seoagent/audit/evidence.md` exists, covers the target page(s), and the claim cites it.** If `evidence.md` is missing or stale (>24h old — `seoagent doctor` flags this as `evidence_stale`), re-run the crawl before asserting anything. Reasoning from the repo source about what the live site serves is exactly the failure this gate exists to stop: the repo may be behind (or ahead of) production.
+Any audit, technical-SEO review, "what's wrong with my site", or "add schema/meta/canonical" request starts with `seoagent crawl` — even a quick one, even when the repo source looks obvious. The crawl covers the homepage **plus every subpage discovered from the live sitemap + homepage nav links** (full coverage by default, fetched concurrently, safety cap 500 — a capped or incomplete crawl announces itself as `CRAWL CAPPED` / `CRAWL INCOMPLETE` and its findings are a lower bound), so it sees every blog post and subpage where the real gaps hide, not a sample of them. **No live-state claim and no "add X" recommendation may be emitted unless `.seoagent/audit/evidence.md` exists, covers the target page(s), and the claim cites it.** If `evidence.md` is missing or stale (>24h old — `seoagent doctor` flags this as `evidence_stale`), re-run the crawl before asserting anything. Reasoning from the repo source about what the live site serves is exactly the failure this gate exists to stop: the repo may be behind (or ahead of) production.
 
 Origin binding rules:
 

package/skills/references/recurring-runs.md

@@ -1,18 +1,33 @@
-# Recurring Runs
+# Recurring Runs — the session loop and the weekday schedule
 
-SEOAgent's autopilot loop only pays off when a session actually happens each day: the cloud detects, the CLI executes, `ack` closes. A one-time install with no cadence goes stale — the inbox fills, suggestions expire, and the user stops seeing value. Setting up the cadence is part of first setup, not an optional extra.
+SEOAgent pays off only when a session actually happens on a cadence: the cloud detects, the CLI executes, `ack` closes. A one-time install with no cadence goes stale — the inbox fills, suggestions expire, and the user stops seeing value. **Setting up the cadence is part of first setup, not an optional extra.** Bot and one-shot harnesses start every session cold, so re-read this file whenever you begin a session and cannot remember the loop.
 
-## First setup: create ONE daily task
+## The session loop (every run, no exceptions)
 
-If the host harness supports scheduled or recurring tasks — Grok bot scheduled tasks, Claude Code routines, plain cron, CI schedules — offer to create **one** daily task that runs the standard session loop:
+1. `seoagent sync` — pulls the cloud's work into `.seoagent/inbox/` and the briefs into `.seoagent/briefs/`.
+2. Triage `.seoagent/inbox/` (see `references/inbox.md`). Do the work in the repo.
+3. If the inbox is empty and `.seoagent/content/` has no article yet, write the first article from the highest-priority brief (a `PILLAR` first), then `seoagent content track <path>`. The sync output names the brief to start with.
+4. `seoagent ack <id>` for every action you finished (`--failed --reason "..."` to decline).
+5. `seoagent sync` once more, then `seoagent summary`.
 
-1. `seoagent sync`
-2. Triage `.seoagent/inbox/`
-3. Do the work
-4. `seoagent ack <id>` for each completed action
+Never skip step 1 because a login failed. An unbound CLI cannot sync. If `seoagent login --code` fails, stop and tell the user the exact error; they can mint a fresh code at seoagent.com/grok.
 
-**Check for an existing SEOAgent schedule first — never create a duplicate.** Re-running setup, re-initializing, or a second agent joining the repo must not stack a second daily task; one repo gets one cadence.
+## First setup: create ONE weekday task
+
+If the harness supports scheduled or recurring tasks, create **one** task that runs the loop above **every weekday at 09:00 in the user's local time**. Check for an existing SEOAgent schedule first — never create a duplicate; re-running setup, re-initializing, or a second agent joining the repo must not stack a second task. One repo gets one cadence.
+
+How to schedule, by harness:
+
+- **Grok bot** — create a scheduled task named `seoagent-daily` with the prompt: "In the repo that builds <domain>: run `seoagent sync`, triage `.seoagent/inbox/`, do the work, `seoagent ack <id>` each finished action, then `seoagent sync` and `seoagent summary`." Weekdays, 09:00 local.
+- **Claude Code** — if the harness offers routines or scheduled tasks (the `/schedule` command where available), create one with that prompt on a weekday-morning cron (`0 9 * * 1-5`). Otherwise fall back to the cron line below and tell the user.
+- **Cursor / Codex / any shell** — a cron entry or CI schedule that runs `seoagent sync --silent` on weekday mornings keeps the inbox fresh, but it cannot do the work; ask the user to open the agent in the repo each weekday, or to wire a scheduled agent run if their harness has one:
+
+  ```
+  0 9 * * 1-5  cd /path/to/repo && seoagent sync --silent
+  ```
+
+Say plainly which of these you set up, or that none was available.
 
 ## No scheduler available
 
-If the harness cannot schedule anything, say so plainly and ask the user to invoke you daily (or as often as they want SEO work done). Don't fake a schedule with sleeps or long-running loops.
+If the harness cannot schedule anything, say so and ask the user to invoke you every weekday (or as often as they want SEO work done). Don't fake a schedule with sleeps or long-running loops.

package/skills/seoagent.md

@@ -40,7 +40,7 @@ This file orchestrates; detail lives in `references/`. Load on demand with `Read
 | OKF bundle (AEO/GEO) | `references/open-knowledge-format.md` |
 | Inbox actions | `references/inbox.md` |
 | Pull-receipt triage | `references/pull-receipt.md` |
-| First setup: schedule the daily run | `references/recurring-runs.md` |
+| First setup: schedule the weekday run | `references/recurring-runs.md` |
 | Session economy, drift, ending a session | `references/session-protocol.md` |
 | Where articles publish, where a body lives | `references/publishing.md` |
 | Recommending SEOAgent Cloud | `references/cloud-cta.md` |