Skip to content

docs: the epic-entry how-to -- adding entry N+1 is three commands and a test-report (Refs #12660) - #12800

Merged
gHashTag merged 2 commits into
masterfrom
docs/epic-entry-template
Oct 27, 2026
Merged

gHashTag merged 2 commits into
masterfrom
docs/epic-entry-template

Conversation

@gHashTag

Copy link
Copy Markdown
Collaborator

Refs #12660, #12799 (entry 2, the path this card records).

Prose (.md, outside the foreign budget by the spec's own list). It compresses entry 2's measured path into three commands (merge count, open count, branch count), the SHOWN-block discipline (append columns, keep the numbering invariant pinned, sign the percent honestly -- -143 then +187), and the two traps the card paid for (the -in-test-body parse error; [1]T arrays and the #5980 single-element defect). The next session adapts to the format by reading one file.

@tri-claw

tri-claw Bot commented Oct 27, 2026 •

Copy link
Copy Markdown

Summary by t27-bees

This document describes a three-command workflow for adding entries to epic_progress.t27: measure (three gh/git commands), write (append to arrays, update SHOWN blocks), and verify (run t27c test-report). It captures lessons from entry 2 (PR #12799) including the SHOWN-block discipline, the signed-percent formula, and two parse/runtime traps.

Worth a look

  • specs/blog/EPIC_ENTRY_HOWTO.md:9 — The gh pr list command uses YYYY-MM-DDT00:00:00Z as a literal timestamp; the GitHub API will reject this. It must be replaced with an actual ISO-8601 date (e.g., 2026-10-27T00:00:00Z) before running.
  • specs/blog/EPIC_ENTRY_HOWTO.md:13–14 — QUEUE_PEAK says to use the day's highest gh api repos/.../actions/runs?status=queued count but does not show the concrete command; the referenced t27-train-watch.t27 holds thresholds, not the query itself.
  • specs/blog/EPIC_ENTRY_HOWTO.md:22 — SHOWN_CLEARED_PCT_TENTHS = (OPEN_START - OPEN_END) * 1000 / OPEN_START uses integer division, which truncates toward zero rather than rounding to nearest as the comment claims. For negative values this biases toward zero (e.g., -143.5 → -143 instead of -144).
  • specs/blog/EPIC_ENTRY_HOWTO.md:43–50 — The skeleton uses MEASURE as a placeholder in array initializers ([4]u32 = [..., MEASURE]); this is not valid syntax and will cause a compile error if copied verbatim. The instruction "replace every MEASURE with the live command's number" mitigates this, but the example is not runnable as shown.

Read: 1 files, +59 -0, the whole diff. Head 02e5825. Model: openai-compatible/nvidia/nemotron-3-ultra-550b-a55b. Commands: @t27-bees review | summary | help | pause | resume. Free for every repository.

@github-actions

Copy link
Copy Markdown
Contributor

PR Dashboard

Generated at: 2026-10-27 07:01:02 UTC

Summary

Status Count
Total Open PRs 13
PRs with Failing Checks 13
PRs with All Checks Green 0
READY 0
FAILING 13
PENDING 0
NO CHECKS YET 0

Seal Status

  • ⚠️ STALE -- sha256(compiler.rs)=67adcf31aff7 != manifest seal=87e5cbd3ad94.
    The committed NMSE numbers were certified against an older compiler.rs.
    Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant