test(planning): probe user_config substitution of an unset default and a saved boolean in a fresh session #9886

Activity

1. kyle-sexton

   added

   needs-humanHuman-in-the-loop required; autonomous sessions must not resolve items carrying this.Human-in-the-loop required; autonomous sessions must not resolve items carrying this.

   on Oct 24, 2026
2. kyle-sexton

   added the

   Task

   issue type on Oct 24, 2026
3. github-actions commented on Oct 24, 2026

   

   github-actionsBot

   on Oct 24, 2026 – with GitHub Actions

   Last edited by melodic-automation-lanes

   Contributor

   More actions

   Proposed work class: work-class: read-only
   Basis: automated intake triage; the lanes App applied this class.

   Size: Small
   Complexity: Low
   Uncertainty: Knowable
   Flags: none
   Blocked by: none

   Understood: this is a probe. In a fresh Claude Code session, with plugin settings saved first, record the Claude Code version. For each of two cases, record whether skill content shows the substituted value or the literal ${user_config...} token. The cases are an unset userConfig key that has a default, and a saved boolean use_emoji_question_markers: true. The results go in a comment. If a case shows a literal token, the comment says so. If the results contradict the "unexpanded means unset" wording in the interview skill, a follow-up issue is filed.

   The probe needs a live claude session, which an unattended lane run can't do. Admission will hold that part. Nothing else looks missing, and I found no likely duplicates among the open issues.

   _{Automated triage, applied by the lanes App.}
4. github-actions

   added

   priority: mediumReal value, no hard deadline; normal backlog flow.Real value, no hard deadline; normal backlog flow.

   on Oct 24, 2026
5. melodic-automation-lanes commented on Oct 24, 2026

   

   melodic-automation-lanesBot

   on Oct 24, 2026

   Contributor

   More actions

   The triage verdict left this issue at status: needs-decision:

   • held by needs-human
   • no lanes App verdict comment shows whether an earlier verdict was flagged
6. melodic-automation-lanes

   added

   status: needs-decisionAwaiting a human or maintainer judgment call.Awaiting a human or maintainer judgment call.

   work-class: read-onlyAudits, research, reports. No repository mutation; tracker and queue writes only.Audits, research, reports. No repository mutation; tracker and queue writes only.

   on Oct 24, 2026
7. kyle-sexton

   added

   host-runNeeds an unattended run on a host; the host runner takes it, not a person or the issue lane.Needs an unattended run on a host; the host runner takes it, not a person or the issue lane.

   status: confirmedA maintainer confirmed the item is real and in scope; ready to be worked.A maintainer confirmed the item is real and in scope; ready to be worked.

   and removed

   needs-humanHuman-in-the-loop required; autonomous sessions must not resolve items carrying this.Human-in-the-loop required; autonomous sessions must not resolve items carrying this.

   on Oct 27, 2026
8. kyle-sexton commented on Oct 28, 2026

   

   kyle-sexton

   on Oct 28, 2026

   MemberAuthor

   More actions

   Host run. I ran fresh claude -p sessions on Claude Code 2.1.312 against a throwaway --plugin-dir plugin, with saved values passed through a one-session --settings file. Result: a saved boolean true substitutes, and an unset key that has a default stays a literal token.

   What ran:

   • Test plugin uprobe. Its userConfig has four keys:
     • unset_str_default: string, default DEFAULT_STR_9886
     • unset_bool_default: boolean, default true
     • saved_bool: boolean, default false
     • saved_str_control: string, the positive control
   • Skill uprobe:show. Its body holds one ${user_config.KEY} line per key.
   • Saved run. --settings held {"pluginConfigs":{"uprobe@inline":{"options":{"saved_bool":true,"saved_str_control":"SAVED_CONTROL_9886"}}}}. The command was claude -p --plugin-dir <plugin> --settings <file> --permission-mode auto --output-format stream-json "Invoke the skill uprobe:show ...". The text below is read from the raw Skill tool result in the stream, not from the model's reply. The session made no Read call.
   • Control run. The same command without --settings.
   • User settings were not modified: the sha256 of the user settings.json matched before and after.

   Results, as the skill content showed them:

   Key	With saved settings	Without settings
   unset_str_default (default DEFAULT_STR_9886)	${user_config.unset_str_default} (literal)	literal
   unset_bool_default (default true)	${user_config.unset_bool_default} (literal)	literal
   saved_bool (saved true)	true	literal
   saved_str_control (saved string)	SAVED_CONTROL_9886	literal

   Against the acceptance criteria:

   1. Version and per-case text: recorded above. (1) An unset key with a default does not substitute into skill content; the literal token arrives. (2) A saved JSON boolean true substitutes as the text true.
   2. Literal token where the doc says a value: the manifest reference ("Reference a saved value") says nothing about either case, so no doc statement is contradicted. One point for the owner: the 2.1.288 changelog implies that defaults substitute for LSP servers, but they do not in skill content. Whether to report that upstream is the owner's call.
   3. Interview wording: plugins/planning/skills/interview/SKILL.md lines 35, 85 and 163 ("unexpanded means unset", "unexpanded or empty means off") match this result: an unexpanded token means no saved value. No follow-up issue was filed. One caveat: a skill must apply a key's own default itself when the token is unexpanded. For a default-true key such as shared_session_open_browser, reading "unexpanded" as "off" would invert that default.

   Target labels: add none; remove host-run.
9. kyle-sexton

   removed

   host-runNeeds an unattended run on a host; the host runner takes it, not a person or the issue lane.Needs an unattended run on a host; the host runner takes it, not a person or the issue lane.

   on Oct 28, 2026
10. kyle-sexton

    closed this as completedin #10675on Oct 28, 2026
11. kyle-sexton

    added a parent issue on Nov 1, 2026

    • fix(planning): correct interview skill docs, shrink required reading, smooth round tooling #6050