diff --git a/skills/firecrawl-scrape/SKILL.md b/skills/firecrawl-scrape/SKILL.md
index 6c6612756c..a6814a8a52 100644
--- a/skills/firecrawl-scrape/SKILL.md
+++ b/skills/firecrawl-scrape/SKILL.md
@@ -51,8 +51,17 @@ firecrawl scrape "https://example.com/pricing" --query "What is the enterprise p
 | `--include-tags <tags>`  | Only include these HTML tags                                     |
 | `--exclude-tags <tags>`  | Exclude these HTML tags                                          |
 | `--redact-pii`           | Redact personally identifiable information from output           |
+| `--max-age <milliseconds>` | Maximum age of cached content. Use `0` to request fresh retrieval |
+| `--json`                  | Emit JSON so metadata such as `metadata.cache` can be inspected  |
 | `-o, --output <path>`    | Output file path                                                 |
 
+## Freshness and cache provenance
+
+- Use `--max-age 0` to request a fresh retrieval path before an action that depends on current state. It skips Firecrawl's index cache and can be slower or more failure-prone than allowing cache.
+- Use `--json` when you need to inspect metadata. An attested Firecrawl-index cache hit includes `metadata.cache.source = "firecrawl-index"` and `metadata.cache.cachedAt`.
+- If `metadata.cache` is absent, do not infer miss, fresh scrape, bypass, or source-page liveness. It only means cache provenance was not attested.
+- For job postings, event pages, inventory, and other stateful records, verify the business object itself before acting: prefer an ATS/source API when available, or check the returned content for active apply buttons, closed labels, canonical IDs, and redirects. HTTP 200 and rendered HTML are not enough.
+
 ## Tips
 
 - **Prefer plain scrape over `--query`.** Scrape to a file, then use `grep`, `head`, or read the markdown directly — you can search and reason over the full content yourself. Use `--query` only when you want a single targeted answer without saving the page (costs 5 extra credits).