Two claims in the roadmap that are not true, both in the v0.11 section - #27
Conversation
|
Preview deployment for your docs. Learn more about Mintlify Previews.
💡 Tip: Enable Automations to automatically generate PRs for you. |
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Advanced Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Found while writing v0.11's build prompts, which would otherwise have inherited both. **The receipt schema count is wrong, for the second time.** The v0.11 line says a chain kept from v0.6 holds four receipt schema versions and names `ctrlrun.receipt/v4` as the schema today. It holds five, and the schema today is `v7`. The sentence already carried a note saying it had been corrected once, for v0.7's bump; v0.10's hop field made it wrong again. Corrected, with the note now saying that twice, because a count of versions in prose goes stale at every release and the argument for reading `receipt.py`'s constants is stronger than the count. **`docs/CONTROL-MAPPING.md` is cited in the present tense and does not exist.** The retention line says it "maps receipt retention to a clause". Line 136 of the same file says it is written only when a design partner asks, which is the truth. The retention debt is real and `docs/postgres.md` is the citation for it, so that half stays and gains the reason the debt is hard: deleting receipts from the middle or the end of the chain is detected as a break by design. Worth separating, because a compliance clause that does not exist is the worst possible input to a milestone's design. `docs/OWASP-AGENTIC-TOP10.md` opens by refusing exactly this shape of claim, and a roadmap that quietly assumes one contradicts it.
15f19b4 to
d79370c
Compare
1 participant
Found while writing v0.11's build prompts, which would otherwise have inherited both. Documentation
only.
1. The receipt schema count is wrong, for the second time
The v0.11 line says a chain kept from v0.6 holds four receipt schema versions, and names
ctrlrun.receipt/v4as the schema today.It holds five, and the schema today is
v7. Probed againstreceipt.py, not counted fromthe prose:
The sentence already carried a note saying it had been corrected once, for v0.7's bump. v0.10's
hopfield made it wrong again. The note now says so twice, and adds the general point: a count ofversions written in prose goes stale at every release, which is the argument for reading the
constants rather than the sentence.
2.
docs/CONTROL-MAPPING.mdis cited in the present tense and does not existThe retention line says it "maps receipt retention to a clause". Line 136 of the same file says
it is "Written only when a design partner asks", which is the truth. The file is nowhere in this
repository and nowhere on disk.
The retention debt itself is real, so that half stays, with
docs/postgres.mdas its citation andthe reason the debt is hard now stated beside it: deleting receipts from the middle or the end of
the chain is detected as a break by design, which is why a retention job cannot simply delete.
Worth separating rather than leaving, because a compliance clause that does not exist is the worst
possible input to a milestone's design.
docs/OWASP-AGENTIC-TOP10.mdopens by refusing exactly thisshape of claim, and a roadmap that quietly assumes one contradicts it. v0.11's prompts now say so
explicitly: do not cite it, do not create it, do not let it shape the item.
Checked
Whole docs suite green against ctrlrun
main. No generated page changes;ROADMAP.mdis the onlyfile touched and it is hand-written.
Stacked on #26, which is why the base is not
main. ctrlrun#191 merged while this was open, somainhere now has a staleneeds_approvalAPI page and anything branched off it inherits thefailure. #26 is the pair that regenerates it. Nothing in this PR depends on that content; the stack
only exists so this one can be green before its turn.
Merge order: #24, then #26, then this.
Not merging.