From 9118bb16e98ed2e19d7430bc7126b3e90ca1e728 Mon Sep 17 00:00:00 2001 From: Daniel JB Clark Date: Sat, 31 Oct 2026 21:31:19 -0400 Subject: [PATCH 1/2] AGENTS: standing rule for most-important work (pedantic, Opus lead, cross-vendor frontier review) --- AGENTS.md | 40 ++++++++++++++++++++++++++++++++++++++++ 1 file changed, 40 insertions(+) diff --git a/AGENTS.md b/AGENTS.md index bb4b420..b02064b 100644 --- a/AGENTS.md +++ b/AGENTS.md @@ -19,6 +19,46 @@ used elsewhere in the ops-djbclark suite (e.g. `site-private/memory/`): directs in the moment — this file doesn't blanket-authorize pushing arbitrary work. +## Standing rule: most important work — be pedantic, pay for the best models (2026-10-31) + +djbclark, 2026-10-31: tendcf and nix2cf are the most important work he is +doing, and getting it right outranks cost. This overrides `model-routing`'s +cost-first default and its "frontier models are scarce" gating (item 1b) **for +these two repos only**; every other repo keeps the default. + +1. **Be pedantic.** Re-read what you changed. Verify every claim by running it, + never by recall. Read the plan or spec section before changing what it + describes. On a doubt, look it up (the CFEngine book, upstream source, the + host itself) instead of assuming. Take the slower path; never narrow scope + or call something verified that you did not run; say what you did not check. +2. **Models.** The lead session runs Opus 5.5 `high` (never Sonnet as lead; + Fable is never a lead). Code slices run on Sonnet 5.5 `high` through + `/bigteam` and climb a tier only on evidence of failure. +3. **Review is cross-vendor, never same-model.** An Opus review of Opus work is + one model with a fresh context, not an independent check. No PR merges on a + Claude-only review. + - Every PR: one frontier review before merge. Order: Codex `gpt-6-astra` + (`acp-dispatch codex --model gpt-6-astra --kind frontier --effort high`), + then Fable `high` (`acp-dispatch claude --model fable --kind frontier + --effort high`), then Gemini 4 Argon (`acp-run agy --model gemini-4-argon + --effort high`; probed working 2026-10-31, own weekly window). + - Design rulings (a `schema_version` bump, the privilege or enforcement scope + on a host, a deploy plan, the unit-writer or consent-diff design): two + independent frontier reviews, astra and Fable; Argon stands in for + whichever of the two is short. A disagreement goes to djbclark, not to the + lead to settle. + - If all three are short, wait for the reset. Never merge on a weaker + review, and never substitute an Opus review. + - The PR body names each reviewer model and effort, quotes the verdict (or + "no findings") and lists what was fixed. +4. **Limits that still hold.** The quota guard and in-flight reservation + (`acp-dispatch` exit 5); no Fable `xhigh`/`max` without djbclark's approval; + one astra slice at a time (four at once emptied Codex's 5-hour window on + 2026-10-18); every slice goes through `acp-dispatch` with `--kind frontier`. +5. **Gemini 4 Carbon** (announced 2026-10-28, not yet accessible to us): recheck + weekly with `aiuse --available --json` and `acp-run agy`; when it is + reachable, adopt it as the Google frontier model in place of Argon. + ## Graft — repo context graph From cd7cb2615aec4205302badd250a68c6ad600a397 Mon Sep 17 00:00:00 2001 From: Daniel JB Clark Date: Sat, 31 Oct 2026 21:48:32 -0400 Subject: [PATCH 2/2] AGENTS: standing rule revised per Fable review (cross-model, short defined, runnable commands, mechanics not overridden) --- AGENTS.md | 92 ++++++++++++++++++++++++++++++++++++------------------- 1 file changed, 60 insertions(+), 32 deletions(-) diff --git a/AGENTS.md b/AGENTS.md index b02064b..d008160 100644 --- a/AGENTS.md +++ b/AGENTS.md @@ -15,49 +15,77 @@ used elsewhere in the ops-djbclark suite (e.g. `site-private/memory/`): - Push immediately after committing. Don't leave it local awaiting a separate "ok to push" prompt. - This exception is narrow: `docs/handoffs/` only. Ordinary code/doc - changes in this repo still go through normal review as the operator - directs in the moment — this file doesn't blanket-authorize pushing - arbitrary work. + changes in this repo go through the standing rule below (a PR with a + cross-model frontier review); this file doesn't blanket-authorize + pushing arbitrary work. ## Standing rule: most important work — be pedantic, pay for the best models (2026-10-31) djbclark, 2026-10-31: tendcf and nix2cf are the most important work he is -doing, and getting it right outranks cost. This overrides `model-routing`'s -cost-first default and its "frontier models are scarce" gating (item 1b) **for -these two repos only**; every other repo keeps the default. +doing, and getting it right outranks cost. This lifts `model-routing` item 1b's +*reason* test for these two repos only: here a frontier review needs no "really +needs a frontier model" case, and a second opinion is reason enough. Every +other repo keeps the default. The mechanics stay (item 4): `acp-dispatch`'s +estimate floors, per-wave caps, in-flight reservation and exit 5 are not +overridden. Never use `route-override`, `--est-pct`, `--min-headroom 0` or +`--no-preflight` to get around them. 1. **Be pedantic.** Re-read what you changed. Verify every claim by running it, never by recall. Read the plan or spec section before changing what it - describes. On a doubt, look it up (the CFEngine book, upstream source, the - host itself) instead of assuming. Take the slower path; never narrow scope - or call something verified that you did not run; say what you did not check. -2. **Models.** The lead session runs Opus 5.5 `high` (never Sonnet as lead; - Fable is never a lead). Code slices run on Sonnet 5.5 `high` through - `/bigteam` and climb a tier only on evidence of failure. -3. **Review is cross-vendor, never same-model.** An Opus review of Opus work is + describes. When in doubt, look it up (the CFEngine book, upstream source, + the host itself) instead of assuming. Take the slower path. Never narrow + scope or call something verified that you did not run, and say what you did + not check. +2. **Models.** Every lead session runs Opus 5.5 `high`, pipeline-only sessions + (merges, waiters) included; `[lead].pipeline_model` does not apply here. + Never Sonnet as a lead, and Fable is never a lead. Code slices run on + Sonnet 5.5 `high` or better (agy's Opus pool is fine) on whichever pool + `route_agent.py plan` picks, through `/bigteam`. Never use a weaker tier + (GLM, Gemini Flash, Cursor Auto, `gpt-6-luna`) in these repos. A slice + climbs a tier only on evidence of failure. +3. **Review is cross-model, never same-model.** An Opus review of Opus work is one model with a fresh context, not an independent check. No PR merges on a - Claude-only review. - - Every PR: one frontier review before merge. Order: Codex `gpt-6-astra` - (`acp-dispatch codex --model gpt-6-astra --kind frontier --effort high`), - then Fable `high` (`acp-dispatch claude --model fable --kind frontier - --effort high`), then Gemini 4 Argon (`acp-run agy --model gemini-4-argon - --effort high`; probed working 2026-10-31, own weekly window). - - Design rulings (a `schema_version` bump, the privilege or enforcement scope - on a host, a deploy plan, the unit-writer or consent-diff design): two - independent frontier reviews, astra and Fable; Argon stands in for - whichever of the two is short. A disagreement goes to djbclark, not to the - lead to settle. + review by the model family that wrote the change. For Opus- or + Sonnet-authored work, astra, Fable or Argon count; for a Codex-authored + slice, Fable or Argon count. + - Every PR, docs-only included (this rule's own PR was reviewed this way), + gets one frontier review before merge. `docs/handoffs/` commits are not + PRs and keep the push-in-place exception above. Try the reviewers in this + order: + 1. astra: `acp-dispatch codex --model gpt-6-astra --kind frontier --effort high --perm read-only --frontier-reason " standing rule: cross-model review" --name NAME -f BRIEF -C WORKTREE` + 2. Fable: `acp-dispatch claude --model fable --kind frontier --effort high --perm read-only --frontier-reason "..." --name NAME -f BRIEF -C WORKTREE` + 3. Gemini 4 Argon: `acp-run agy --model gemini-4-argon --effort high -C WORKTREE -f BRIEF`, run with read-only instructions in the brief. Argon was probed working on 2026-10-31 and has its own weekly window. Use `acp-run` until djbclark-ade's Argon frontier wiring lands; until then, record the model, effort and `aiuse` reading in the PR body yourself. + - A pool is **short** when `acp-dispatch` answers exit 5 (`unavailable`: + wont-fit, inflight, headroom or preflight), when `aiuse --available --json` + lists it `usable_now: false`, or, for Argon, when `acp-run agy` fails on + quota. Short means move to the next model in the order, never force it. + Say in the PR body which model was short and why. As of 2026-10-31 astra + is short: Codex has 45% left, below the 60% `codex:frontier` estimate + floor, so Fable takes the first seat. + - Design rulings (a `schema_version` bump, the privilege or enforcement + scope on a host, a deploy plan, the unit-writer or consent-diff design) + get two independent frontier reviews, astra and Fable. Argon stands in + for one of them and never counts twice. With both astra and Fable short, + take the Argon review now and wait for a reset for the second. A + disagreement goes to djbclark, not to the lead to settle. - If all three are short, wait for the reset. Never merge on a weaker - review, and never substitute an Opus review. + review, and never substitute an Opus review. A GitHub app review (for + example the chatgpt-codex-connector bot) is not the frontier review. + - The `/bigteam` adversary round (omp Opus) still runs where bigteam calls + for it. It is additional and never counts as the frontier review. - The PR body names each reviewer model and effort, quotes the verdict (or "no findings") and lists what was fixed. -4. **Limits that still hold.** The quota guard and in-flight reservation - (`acp-dispatch` exit 5); no Fable `xhigh`/`max` without djbclark's approval; - one astra slice at a time (four at once emptied Codex's 5-hour window on - 2026-10-18); every slice goes through `acp-dispatch` with `--kind frontier`. -5. **Gemini 4 Carbon** (announced 2026-10-28, not yet accessible to us): recheck - weekly with `aiuse --available --json` and `acp-run agy`; when it is - reachable, adopt it as the Google frontier model in place of Argon. +4. **Limits that still hold.** The quota guard, estimate floors and in-flight + reservation (`acp-dispatch` exit 5); per-wave frontier caps + (`config/gates.toml [frontier]`); no Fable `xhigh`/`max` without djbclark's + approval; one astra slice at a time (four at once emptied Codex's 5-hour + window on 2026-10-18). Every astra or Fable review goes through + `acp-dispatch --kind frontier`, never a hand-run `acp-run`. Argon is the + one exception, for the reason given in item 3. +5. **Gemini 4 Carbon** (announced 2026-10-28, not yet reachable by us): + recheck weekly with `acp-run agy --info --model gemini-4-carbon` and + `aiuse --available --json`. When it is reachable, adopt it as the Google + frontier model in place of Argon. ## Graft — repo context graph