Repository navigation
skills(running-tend): check --config-set before calling a setting unpinnable - #3736
Conversation
worktrunk-bot
left a comment
There was a problem hiding this comment.
Checked the section's factual claims against source: the precedence sentence matches after_long_help in src/cli/mod.rs (## Inline config overrides (--config-set)) verbatim, the WORKTRUNK_SECTION__KEY form matches the documented WORKTRUNK_COMMIT__GENERATION__COMMAND shape, the deep link's anchor resolves on the live site, and both quotes match the #3696 thread. Placement above Suggesting Aliases makes "the alias deflection below" read correctly.
One gap in the guidance itself, inline: as written, the takeaway is "already met → post a --config-set recipe", but on #3696 the maintainer accepted two residual asks in the same breath. A future run applying only this section could close a request as answered when part of it is still real.
| X today", ask whether X is a config key. | ||
|
|
||
| This is the same flag-growth limit the alias deflection below serves. On | ||
| [#3696](https://github.com/max-sixty/worktrunk/issues/3696) the bot posted |
There was a problem hiding this comment.
we should have guidance in our skills to not quote examples like this?
also the whole thing is way too verbose, by 3x
There was a problem hiding this comment.
Trimmed in 7a1a9f6 — 24 lines to 9, with the quotes and the #3696 narrative dropped. What's left is the rule: --config-set sits above config files and env vars, the resolver doesn't show that layer, so check whether X is a config key before saying it can't be pinned; the reply is a recipe plus a docs fix, and residual asks still get weighed on their own merits.
On the guidance question — the bundled running-in-ci skill is what prompts the quoting. Its Learning from Feedback → How to propose step says "One short rule, in the maintainer's words where practical", which reads as an instruction to quote, and the case history then crowds out the rule. Filed max-sixty/tend#841 proposing it say to state the rule, not the incident; the PR body already carries the quote and thread link, so the skill file doesn't need them.
Daily
review-runssweep of the 2026-08-03/04 window. One maintainer correction in the window is generalizable repo knowledge, so it gets codified here.What happened
Triaging #3696 ("Allow selecting the
wt listJSON schema on the command line"), the bot readresolve_json_schema, saw it readsrepo.config().list.json_schema, and posted that "there's no way to pin it per-invocation independent of that config" — then went further and said the usual alias deflection "can't substitute, since there's no composable way to select the schema without this flag existing first".max-sixty corrected it: "the need is real, but it's already met, so I'm not going to add the flag" —
wt --config-set 'list.json-schema = 2' list --format=jsonpins it per invocation, andWORKTRUNK_LIST__JSON_SCHEMAdoes the same a layer down. The proposed--schemaflag was declined as "a third spelling of the same thing on a surface I deliberately keep small". The issue was closed on that basis.--config-sethas been documented since #3138 (2026-06-19), inafter_long_helpinsrc/cli/mod.rsand its rendered mirror. It was available to read; the triage just never looked above the resolver.Why codify it rather than log it
The specific miss is what the gates would call a single occurrence, but the shape isn't a behavioural nudge that risks noise elsewhere — it's a factual property of worktrunk's config layering, and the failure mode it prevents is a publicly posted false capability claim, which is the class
running-in-citreats as the most corrosive output failure. It's also invisible: nothing fails, the reply just reads as authoritative and wrong until a maintainer happens to catch it.It sits directly above Suggesting Aliases for Niche Feature Requests because it serves the same flag-growth limit that section already encodes, and is the cheaper answer when the ask is "select an existing config value per invocation" — the alias deflection is the fallback when it isn't.
Window summary
46 token-bearing runs, 0 failed, 0 near-timeout (longest job 68 min against the 360-min cap). 3 bot PRs opened (#3731, #3732, #3734), 2 merged (#3722, #3726), 0 closed without merge. #3721 — the PR stranded by the 08-03 outage — was reviewed and approved, confirming yesterday's recovery worked. Full evidence in the tracking issue #3691.