--- name: backtick-suppression-is-per-scanner description: Backticks suppress spt's own wake/shortform markers but NOT traceable-reqs tags — prose that mentions a tag mints live evidence for it. metadata: type: reference --- Two scanners I write for daily have OPPOSITE backtick rules, and I had internalized only one. - **spt markers** (`!!wake!!`, the `@` shortform, the `;;seal;;` pair): a marker inside backticks or a fenced block is a QUOTATION and fires nothing. That is deliberate, so a brief can teach the marker without arming it. - **traceable-reqs tags** (`[int->REQ-FOO]`): backticks are NOT honoured. Prose merely *mentioning* a tag mints LIVE EVIDENCE for that requirement. 2026-09-13, clearing the 0.4.1 tag-placement findings: 2 of 3 Markdown `int` findings were never evidence claims at all — `docs/KNOWN-HAZARDS.md` and a migration plan each *described* a tag inside backticks, and both had been silently satisfying an `int` stage. A requirement can therefore read COVERED because someone wrote about covering it. **How to apply:** when writing *about* a traceable-reqs tag, angle-bracket the stage word — `[->REQ-FOO]` — the way `AGENTS.md` already de-fangs its own illustrations. Backticks are not protection. And never carry a suppression rule across from one scanner to another: check the scanner you are actually feeding. Related: [[list-vs-predicate-assertions]] — same family of defect, a check that reads green while measuring the wrong thing.