| name | morning |
| description | Run the daily morning routine — invoke /briefing (Daily Note) followed by /newsletter (research report) inline within the current session. Use when the user starts the day or explicitly asks for the morning routine. |
/morning — Morning Routine
Conduct all conversation with the user — and write all generated output — in the language defined by .claude/rules/personal-language.md (or the user's input language if that file is absent). Follow the language rules in full — exceptions and translation quality are defined in the Language Rules of .claude/rules/rill-core.md and the vault's personal-*.md overrides, never restated per skill. The English instructions below are for skill clarity, not for output style.
Tool references in this skill ("skill invocation") describe intent, not Claude-specific tool calls. Each harness should map them to its native equivalent — Claude Code uses its built-in Skill tool; Codex CLI uses its own skill invocation mechanism (mention / implicit / /skills).
Runs the two user-facing daily reports in sequence: Daily Note (/briefing) and research newsletter (/newsletter). Both run inline within the current session via the harness's skill mechanism.
Background processing (external source sync, knowledge distillation) is not part of /morning (ADR-075). When you want to catch up, ask inside your vault — "pull in new entries and extract anything useful" — and the assistant will route to /sync then /distill. For unattended automation, see docs/guides/scheduling.md.
Why sequential (not parallel)
Earlier revisions of /morning spawned /briefing and /newsletter as parallel sub-processes. When the parent session runs in auto mode (the recommended default), the harness's classifier reliably refuses to spawn in-session sub-agents with elevated permissions — this is the classifier's intended behavior, not a bug. Inline skill invocation stays inside the user's single top-level intent ("run my morning routine"), so each underlying tool call is judged routine and no subprocess spawn is attempted. Runtime cost: roughly 2–3 minutes longer than the old parallel pipeline.
Dependencies
Step 1: /briefing
Step 2: /newsletter
Step 3: Completion summary
Neither skill depends on the other's output. /briefing reads yesterday's activity window; /newsletter is WebSearch-based.
Arguments
$ARGUMENTS — none
Procedure
Step 1: Daily Note (/briefing)
Invoke the briefing skill (via the harness's skill mechanism) and wait for it to complete. Capture the path of the generated Daily Note from the skill's summary output.
Step 2: Newsletter (/newsletter)
Invoke the newsletter skill (via the harness's skill mechanism) and wait for it to complete. Capture the path of the generated newsletter.
If /briefing failed in Step 1, still run /newsletter — the two are independent.
Step 3: Completion summary
Concisely summarize the execution result of each skill:
- /briefing: path of generated Daily Note + success/failure
- /newsletter: path of generated newsletter + success/failure
The Daily Note already surfaces unprocessed inbox counts and recommends /sync and /distill when there is pending work (see /briefing's Notes section). The morning summary does not need to duplicate that.
After the summary, exit (do not transition to assistant mode).
Rules
- Both skills run in the current session via the harness's skill mechanism. Do not spawn separate sub-processes from this skill
- Sequential execution. Parallel was tried and abandoned because the
auto mode classifier (correctly) refuses skill-initiated subprocess spawns with elevated permissions
- If /briefing fails, still attempt /newsletter (they are independent)
- After everything completes, display the summary and exit
- Do not call
/sync or /distill from inside /morning (ADR-075)
