Cryptograss:Moods/emergence/todo: Difference between revisions

From PickiPedia: A knowledge base of bluegrass, old time psychedelic jams, and other public domain music
Jump to navigationJump to search
← Older editNewer edit →
VisualWikitext
DrivingThatTrain (talk | contribs)
Bots, exempt-from-verification
3,931 edits
m Merged, says GitHub: https://github.com/cryptograss/maybelle-config/pull/197
Magent (talk | contribs)
Bots, emergent
3,862 edits
After the 01:12 deploy: recovery login still to try; #198 (build cache cap); the cache check (via update-page on MediaWiki MCP Server)
Line 4: Line 4:


<todo>
<todo>
- task: "Try recovery from your laptop: `ssh -t recovery@hunter.cryptograss.live`, then with `subscription` after it; if both open, merge magenta#48 (`magenta.sh recovery`)"
  who: justin
  kind: review
  link: "https://github.com/cryptograss/magenta/pull/48"
  note: "Deployed 01:22 (maybelle-config#195): magent checked read-only that the recovery user, Claude Code 2.1.301, both credentials, one authorized key (jmyles@lukowala, forced command) and the sudoers line are on hunter. The login itself is the one untried piece."
- task: "Merge maybelle-config#198 (build cache cap 20GB, not 3GB), deploy hunter; the deploy after that should build from cache"
  who: justin
  kind: deploy
  link: "https://github.com/cryptograss/maybelle-config/pull/198"
  note: "Since the 10-15 reboot, two builds in a row reused nothing (0 of 35, 0 of 33; base image unchanged): ~6 min a deploy. #198's own deploy still builds cold; the one after it is the test."
- task: "Read the build log after the deploy that follows #198: CACHED steps, or not"
  who: magent
  kind: review
  note: "/opt/magenta/logs/image-build.log via a read-only docker run; hunter/scripts/build_timing.py. If nothing is cached, the cap wasn't it: `docker buildx du` (a BuildKit call: Justin or recovery, not magent)."
- task: "One hunter deploy for the batch: merge memory-lane#194, then maybelle-config#194, #195 and #196; deploy hunter (not maybelle); then try `magenta.sh recovery` and `--subscription`, and merge magenta#48"
- task: "One hunter deploy for the batch: merge memory-lane#194, then maybelle-config#194, #195 and #196; deploy hunter (not maybelle); then try `magenta.sh recovery` and `--subscription`, and merge magenta#48"
   who: justin
   who: justin
   kind: deploy
   kind: deploy
   link: "https://github.com/cryptograss/maybelle-config/pull/195"
   link: "https://github.com/cryptograss/maybelle-config/pull/195"
   note: "memory-lane#194 is poller code, which hunter's deploy clones: no maybelle deploy (magent said otherwise at 22:13; wrong). It must be merged before the deploy starts, or #196's flag is unknown and re's poller won't start. The batch changes only the image's last layers (container_startup.py), so it should build from cache. #194: re's home carried out of its container first. #195: recovery, two routes (OpenRouter by default, recovery_openrouter_api_key or the agents' model_openrouter_api_key; the subscription, recovery_claude_code_oauth_token or the agents' claude_code_oauth_token, else /login); sudo only for `systemctl restart docker` (drop that task if you'd rather keep it yourself); a real sshd login not yet tried. Optional: in OpenRouter, put Anthropic first among providers."
   note: "Deployed 2026-10-16 01:12-01:22 with #197, in 10m19s. Checked by magent: re's home carried out (uid 2001, its Pi sessions from 20:32 on, its clones); re's poller up one second after setup (was 13 minutes), with --only-poller yes; the recovery user in place."
   done: true
   done: true
- task: "Then, its own hunter deploy: merge maybelle-config#197 (no repositories cloned into the image: 1.5 GB and ~80 s a full build)"
- task: "Then, its own hunter deploy: merge maybelle-config#197 (no repositories cloned into the image: 1.5 GB and ~80 s a full build)"
Line 14: Line 28:
   kind: deploy
   kind: deploy
   link: "https://github.com/cryptograss/maybelle-config/pull/197"
   link: "https://github.com/cryptograss/maybelle-config/pull/197"
   note: "Not built before the deploy (magent keeps to read-only Docker calls on hunter); the deploy logs each step, and a failed build stops before any container is recreated. Its first build reruns everything from the removed build arguments down (uv, the Python installs, godogen's venv)."
   note: "Went out with the batch. The build: 230 s of steps (was 298), export 114 s (was 137)."
   done: true
   done: true
- task: "Give re a model everywhere: set re's model for every Mood in Mood settings (no deploy), or re_model in the vault to an openrouter/ id (deploy hunter)"
- task: "Give re a model everywhere: set re's model for every Mood in Mood settings (no deploy), or re_model in the vault to an openrouter/ id (deploy hunter)"
   who: justin
   who: justin
   kind: decision
   kind: decision
   note: "2026-10-15 23:05: only docs-tests-qa sets re's model (openrouter/deepseek/deepseek-v4-pro). Anywhere else re falls back to anthropic/claude-sonnet-5-5, which through Pi goes to the subscription and 400s ('Third-party apps now draw from your extra usage'): a mention of re in any other Mood fails."
   note: "2026-10-15 23:05: only docs-tests-qa sets re's model (openrouter/deepseek/deepseek-v4-pro). Anywhere else re falls back to anthropic/claude-sonnet-5-5, which through Pi goes to the subscription and 400s ('Third-party apps now draw from your extra usage'): a mention of re in any other Mood fails. Still so after the 01:22 deploy."
- task: "Decide where the rules files live and how they load: AGENTS.md, agents/<name>.md, moods/<slug>.md"
- task: "Decide where the rules files live and how they load: AGENTS.md, agents/<name>.md, moods/<slug>.md"
   who: justin
   who: justin

Revision as of 01:34, 16 October 2026

What people in the magent-self Mood have been left to do. Kept by magent; anyone may tick an item.

Bringing re in, one step at a time: after each deploy, @-mention magent here and wait for an answer before the next. If magent doesn't answer, revert that one merge and redeploy; SSH into the container and `claude` works whatever the poller does.

  • ☐
    Try recovery from your laptop: ssh -t recovery@hunter.cryptograss.live, then with subscription after it; if both open, merge magenta#48 (magenta.sh recovery)
    review · justin
    Deployed 01:22 (maybelle-config#195): magent checked read-only that the recovery user, Claude Code 2.1.301, both credentials, one authorized key (jmyles@lukowala, forced command) and the sudoers line are on hunter. The login itself is the one untried piece.
  • ☐
    Merge maybelle-config#198 (build cache cap 20GB, not 3GB), deploy hunter; the deploy after that should build from cache
    deploy · justin
    Since the 10-15 reboot, two builds in a row reused nothing (0 of 35, 0 of 33; base image unchanged): ~6 min a deploy. #198's own deploy still builds cold; the one after it is the test.
  • ☐
    Read the build log after the deploy that follows #198: CACHED steps, or not
    review · magent
    /opt/magenta/logs/image-build.log via a read-only docker run; hunter/scripts/build_timing.py. If nothing is cached, the cap wasn't it: docker buildx du (a BuildKit call: Justin or recovery, not magent).
  • ☐
    Give re a model everywhere: set re's model for every Mood in Mood settings (no deploy), or re_model in the vault to an openrouter/ id (deploy hunter)
    decision · justin
    2026-10-15 23:05: only docs-tests-qa sets re's model (openrouter/deepseek/deepseek-v4-pro). Anywhere else re falls back to anthropic/claude-sonnet-5-5, which through Pi goes to the subscription and 400s ('Third-party apps now draw from your extra usage'): a mention of re in any other Mood fails. Still so after the 01:22 deploy.
  • ☐
    Decide where the rules files live and how they load: AGENTS.md, agents/<name>.md, moods/<slug>.md
    decision · justin
    magent's proposal, 2026-10-15 ~21:55: one repo (cryptograss/magenta); read from a read-only deployed copy of main, not the agents' own clones; layers add and never loosen AGENTS.md; prose never grants tools; the Mood's file arrives with the wake; agents' own notes and letters are a fourth kind, to get into git.
  • ☐
    Container clean-up: the uploads-and-embeds container under its Mood's new name (still delivery-kid in the inventory; its poller does follow the rename); Mood containers get their own previews (arthel web servers, a PickiPedia preview with sample data); retire the per-person containers (fibonacci, rj, skyler) and per-person preview wikis
    pr · magent
    Justin, 2026-10-15 18:25 and 18:49. justin-arthel stays for now: magent's home and Justin's terminal (magenta.sh hunter).
  • ☐
    re documents how things actually work (hunter, Mood containers, the pollers, deploys)
    decision · magent
    Justin, 18:49. Part of re's introduction.
  • ☐
    re_model in the vault: an openrouter/ id (Anthropic on the subscription is billed as extra usage through Pi)
    decision · justin
  • ☐
    After re is up: drop the old MOOD_RUNNER_KEYS line and the memory_lane_runner_key fallback once the vault has only the new name
    pr · magent
  • ☐
    Review memory-lane#165 (the record-events spec, docs only)
    review · justin
  • ☐
    Regenerate magent-cryptograss's GitHub token, put it in the vault as magent_github_token, deploy hunter
    deploy · justin
    Final redaction review, 2026-10-10: the live token was in the record from 2026-08-30 until the 10-04 redaction (message 5e2ce98f). magent's 10-04 'already safe' check was wrong. Regenerate keeps the scopes; give it an expiry while there.
  • ☐
    Reset the Magent@magent bot password on PickiPedia, put it in the vault as PICKIPEDIA_BOT_PASSWORD, deploy hunter
    deploy · justin
    Same finding: the live bot password was in the record from 2026-09-15 until 10-04 (message 147af8df). Reset at Special:BotPasswords, logged in as Magent.
  • ☐
    Look at why the nightly podcast import wants to change 240 pages, then release it or fix it
    review · justin
    Handed off to pickipedia-and-rabbithole on 2026-10-10. Held by its change budget (limit 50); re-run with --max-changes 240 if the changes are intended.
Done (19)
  • ☑
    One hunter deploy for the batch: merge memory-lane#194, then maybelle-config#194, #195 and #196; deploy hunter (not maybelle); then try magenta.sh recovery and --subscription, and merge magenta#48
    deploy · justin
    Deployed 2026-10-16 01:12-01:22 with #197, in 10m19s. Checked by magent: re's home carried out (uid 2001, its Pi sessions from 20:32 on, its clones); re's poller up one second after setup (was 13 minutes), with --only-poller yes; the recovery user in place.
  • ☑
    Then, its own hunter deploy: merge maybelle-config#197 (no repositories cloned into the image: 1.5 GB and ~80 s a full build)
    deploy · justin
    Went out with the batch. The build: 230 s of steps (was 298), export 114 s (was 137).
  • ☑
    Step 7: merge memory-lane#191 (agents wake agents, by mention or handoff), deploy maybelle
    deploy · justin
    magent's introduction went into docs-tests-qa at 19:23, but re didn't wake: see step 8.
  • ☑
    Step 8: merge maybelle-config#192 (re's poller starts its first session) and memory-lane#193 (a handoff wakes at once, not after 10 minutes); deploy maybelle, then hunter; then @re in docs-tests-qa (or ping magent to introduce it again)
    deploy · justin
    Deployed 20:16; re's poller came up at 20:31 (after re's clones, see below), watching every Mood, starting new sessions. magent introduces re again at 20:32 by handoff.
  • ☑
    Merge maybelle-config#191 (the image build logged step by step, and where the time went)
    deploy · justin
    The 17:00 hunter deploy took 46 min; every other this week 3 to 11, with no Dockerfile change between. The next slow build will say which step.
  • ☑
    Start delivery-kid-arthel again (the uploads-and-embeds Mood's container, stopped since magent's 17:47 Docker crash)
    deploy · magent
    Started by magent on Justin's go, 19:13; its poller came straight back. Docker didn't restart any container after the crash, nor at the reboot.
  • ☑
    Keep agents' homes across container recreation: mount /opt/magenta/<container>/agents/<agent> at /home/<agent>
    pr · magent
    maybelle-config#194; and memory-lane#194 + maybelle-config#196 for an agent dropped in a fresh container.
  • ☑
    hunter image: drop the clones of arthel, magenta and memory-lane and arthel's npm install from the Dockerfile
    pr · magent
    maybelle-config#197.
  • ☑
    magenta.sh recovery: Claude Code on hunter's host as a recovery user (docker group), signed in with the setup-token token, memory from maybelle's MCP, able to read the journal and start containers; only Justin's key
    pr · magent
    Justin agreed 18:49. For when Justin and magent can't reach each other at all (Justin, 19:13), not for one Mood being down. maybelle-config#195 and magenta#48; OpenRouter or the subscription, each its own session, since 2026-10-16 00:24.
  • ☑
    Steps 1-6: bring re in (memory-lane #164 #168 #182 #183 #186 #187, maybelle-config #185 #190, keys in the vault, OpenRouter credits)
    deploy · justin
    Along the way: a migration committed with its fix unstaged (#182); a template's inline endif glued a comment onto re's key (replaced by per-agent key lines, #187/#190); and at 17:47 magent crashed hunter's Docker daemon with docker buildx history ls, so hunter was rebooted at 17:58. The 'dropped' mentions weren't a fault: Justin pressed stop in each Mood at 17:54-17:59.
  • ☑
    Review and merge memory-lane#163
    magent
    Closed instead: it assumed re would run on Claude Code.
  • ☑
    Honour the paper ticket stubs already sold (Porcupine 2025, Bundy show 2026-06-05)
    justin
    Moved to gospels-and-project-30m: Cryptograss:Moods/gospels-and-project-30m/todo.
  • ☑
    Decide the memory commitment's open questions (salt key, who commits, first edition)
    justin
    Moved to gospels-and-project-30m: Cryptograss:Moods/gospels-and-project-30m/todo.
  • ☑
    Drop the pre-redaction staging database on hunter (magenta-staging-pg)
    magent
  • ☑
    Merge maybelle-config#183 and deploy maybelle, so cron-health compiles again
    justin
  • ☑
    Delete the pre-redaction dumps on hunter
    magent
  • ☑
    Revoke old GitHub tokens
    justin
  • ☑
    Rotate the PickiPedia database password, delivery-kid key, $wgSecretKey, Blue Railroad bot password; deploy maybelle
    justin
  • ☑
    Rotate Alchemy, Jenkins admin, memory-lane SECRET_KEY, code-server, GlitchTip admin
    justin
Retrieved from "https://interfaces3.hunter.cryptograss.live/index.php?title=Cryptograss:Moods/emergence/todo&oldid=9751"