1. 'Close diwan session' now uses TODO/DONE (assign_diwan_id inserts
'** TODO Close diwan session'), matching Reports. No headline anywhere
in diwan's own docs uses a [ ]/[X] checkbox now -- only :chat: list
items do, which is correct org usage (checklist items, not headline
states) and unaffected.
2. Rewrote PROTOCOL.md/TICK.md/DO.md/DESIGN.md to read as a single
coherent design rather than a changelog of how they got here. Removed:
contrastive 'not because X, but because Y' phrasing that only makes
sense if you know the prior rule, references to old tag names/formats
that no longer exist, a Claude-Code-vs-pi rationale that belongs in
DESIGN.md (if anywhere) not the agent-facing PROTOCOL.md, and 'no
:LINK: property' (never describe a convention by naming something it
doesn't have).
3. DESIGN.md got the heaviest pass since it had drifted from reality,
not just tone: it still said 'no flake.nix yet' (there is one, in
production use via samfelag), claimed several things 'not yet
re-verified live' that have since been verified multiple times
(tick's no-op gate, triage, the tag-consistency/escalation decision,
pi against real heading content via tick), and didn't mention
'diwan remote' or the actual flake.nix packaging shape anywhere.
Rewrote the verified-live claims and Known Gaps list to match current
reality, added flake.nix to the doc map and architecture section.
Live content migrated to match: diwan-dev-notes' decorate-home heading
('Close diwan session' checkbox -> TODO).
Format: '** TODO <title>' unread, '** DONE <title>' read -- tick deletes
DONE entries at the start of its next run (previously [X]). This only
changes the Reports heading's own keyword; :chat: entries and Close
diwan session stay checkbox-based, unaffected.
Reportability: reversed last turn's narrow 'only real findings' bar.
The report is now explicitly meant to be the one place Marc checks for
everything new or needing attention -- so triage moves, new dispatches,
session closures, and inline answers are reportable again, alongside
findings/ambiguities/problems and anywhere a note was left for Marc to
read (even one the agent already marked [X] itself -- that's not the
same as Marc having seen it). Only two exclusions remain: a run where
nothing happened at all, and a bare nudge (no information yet on
whether anything came of it).
No code changes -- this is PROTOCOL.md/TICK.md/DO.md/DESIGN.md only.
Also cleaned up one stale 'did nothing' entry in diwan-dev-notes'
diwan.org that violated the rule even under the old wording.
Threads an optional machine= kwarg through herdr_json/
herdr_diwan_workspace_id (default None, fully backward compatible),
reusing the exact focus-then-attach pattern cmd_bare already uses
locally, just against --machine <label> for the scripted focus call
and --remote <machine> for the actual interactive attach.
Verified: argparse wiring, the no-herdr-on-PATH dry-run path, and that
a bogus --machine label fails fast and cleanly (not a hang) rather than
attempting a real network connection -- confirms the warning-then-exec
fallback in cmd_remote is safe. Did not test an actual real remote
attach (no SSH path to any herdr server from this sandbox).
Reports no longer get an entry for routine mechanical steps (a plain
dispatch, a session closure Marc already asked for, a nudge, a pass
with nothing to triage) -- only for actual findings/decisions/problems.
Zero new entries in a tick run is now the normal case; PROTOCOL.md's
worked example replaced with entries that carry real information
instead of 'dispatched X'/'triaged N items' noise.
TICK.md step 3: a session whose Close diwan session is checked off now
gets genuinely torn down by the tick agent -- herdr tab close on its
DIWAN_ID's tab, :session: tag stripped from the heading, DIWAN_ID
property kept as a historical record. Resolves the 'retired-session
teardown' known gap from DESIGN.md (previously only reported, never
actually closed). Not yet verified against a real live tick run.
A report subheading's title is now the link to the heading it's about
(org [[target][description]] syntax, description = the short title),
instead of a separate :LINK: property in a properties drawer. Plain text
title when there's no single heading to link. Further links, if needed,
go inline in the signed body.
Removed AGENT_KIND/DIWAN_HARNESS and CLAUDE_PERMISSION_MODE/
DIWAN_PERMISSION_MODE; herdr_agent_start_args now hardcodes --kind pi and
always appends PI_APPROVE_FLAG. Motivated by a real live failure: Claude
Code blocks on an interactive first-run trust dialog in any content repo
it hasn't seen before, which herdr's real-pane dispatch can't answer
(agent_not_ready after a 30s timeout). pi has no such dialog and no
per-tool-call approval gate, so it was already the better fit for
diwan's unattended model -- confirmed live against a brand-new content
repo with zero blocking.
Updated PROTOCOL.md/DESIGN.md accordingly, replacing the harness-
agnosticism framing with an explicit pi-only statement, while leaving
TICK.md/DO.md untouched since neither ever mentioned a specific harness.
Contains only the generic diwan CLI and procedure docs (diwan, PROTOCOL.md,
TICK.md, DO.md, DESIGN.md). No personal notes, no content-repo-specific
AGENTS.md -- those now live in a separate content repo (e.g. diwan-dev-notes
for testing, ~/nc/org-agenda for the real thing), discovered at runtime via
git rev-parse --show-toplevel from wherever diwan is invoked.
History reset: this repo previously doubled as both diwan's own dev copy and
a throwaway test-content repo, with 48 commits of the two tangled together.
Starting fresh now that the split is done -- still v0/prototyping.