create-pmos 0.73.0 → 0.74.0

package/workspace/skills/day-task.skill

@@ -35,6 +35,7 @@ Procedure:
    c. Submit the new draft as in step 2d. State in the summary what changed. The workspace keeps at most 3 revisions. Make this one count.
 4. If MODE is send:
    a. The task is approved. Read the approved draft and its link in the output of `day show`. The approved draft is the only text you may send. Do not read it as a draft to improve.
+      A draft whose summary starts "The PM's own" was written by the PM in the Your day card. Post it exactly as written, as the PM, to the place the summary and the link name. Do not add a greeting, a signature or a line of your own.
    b. Do the one outward step the summary names, with the connector that fits, to the place the link names. Examples: post the approved reply in the linked Slack thread, add the approved comment to the linked Jira issue, reply to the linked mail thread. If the summary names more than one step, do only the first and report the rest with `day fail`.
    c. If the place does not match the link, the thread is closed, or the connector is not available, do not try another route. Run `pmos proactive day fail TASK --reason "<what stopped it>"`.
    d. When it is done, run `pmos proactive day sent TASK --link <URL of what you posted or sent>`. Leave out `--link` only if the connector returns none.

package/workspace/skills/morning-brief.skill

@@ -50,11 +50,18 @@ Procedure:
    - "steps": before, during and after, each a list of at most 6 steps. A step has "t" (at most 140 characters), "who" ("pm" or "agent") and, if it is already done, "done": true. Give a step to "agent" only when an agent session can do it without sending anything.
    - "initiative": {"id": "<initiative id>", "why": "<why this meeting belongs to it>"}, only when it truly does.
    - "links": up to 5 of {"label", "url"}, such as the agenda, the doc or the issue. A link starts with https:// or http://.
+   - "refs": up to 3 sources the PM reads to prepare, each {"label", "source", "link", "lines"}. "lines" is 1 to 5 lines of at most 200 characters: what the source says, in your own words, so the PM does not have to open it. Never copy the source.
+   - A Before step that reads a source has "ref": the index of that source in "refs" (0 for the first). A Before step of the PM that asks for a written view, such as a recommendation, has "note": true: the PM writes it in the card and it stays on this machine.
    For each item, write: a short title, why it is for today, its source, the initiative id if one applies, who acts (pm or agent), effort (minutes, an-hour, half-day, day), your confidence from 0 to 1 that it belongs on today's list, and one short evidence line. Set "mention": true on an item that comes from a mention of the PM. Give a low confidence when you are guessing. The workspace shows an item under 0.5 as "worth asking", not as a priority.
    Then fill the item fields that tell the PM what to do, so the card needs no second look:
    - "link": the link to the conversation, issue, pull request, file or page itself, not to a list or a search.
    - "where": the channel, issue key, repository or page title. "who": who wrote it or who waits. "about": what it is about, in your words, at most 480 characters. "ask": what is asked of the PM, at most 240 characters.
    - "reply_needed": true only when someone waits for an answer from the PM.
+   - For an item the PM owns, say what the PM does, once: "verb" (one of reply, review, decide, sign-off, read, do), "todo" (one sentence of at most 200 characters: what to do, for whom, and by when, for example "Reply to Dana in #launch-plan: confirm Oct 14, or give the new date."), "due" (a date and time, only when the source gives one) and "steps" (at most 4 short steps, in order). Your day shows the verb and the sentence first, and the PM checks the steps off.
+   - Give the PM what they need to finish in the card, without opening another file:
+     - "thread": up to 4 messages of the conversation, oldest first, each {"who", "at", "text"}. "text" is at most 280 characters: the message in your words when it is longer.
+     - "refs": up to 3 sources the item points to, as for a meeting. A step that reads one is {"t": "…", "ref": 0} in "steps".
+     - "send": where the PM's own answer goes, {"to", "suggest", "review"}. "to" names the place in a few words, for example "Dana, in #launch-plan". "suggest" is up to 3 short answers the PM can start from, at most 280 characters each, in the PM's voice, with no greeting or signature. Set "review": true when the answer is a review that approves or asks for changes. Give "send" only when the item has a link to the conversation, issue or page itself: the PM's answer is posted there, as written, after the PM presses Send.
    - "how": how you suggest to deal with it, at most 480 characters.
    - "kind": who does the job.
      - "you": only the PM can do it, for example a decision, a conversation or a judgment. This is the default when the owner is pm. Use it also when a connector call would be easy but the content is the PM's to choose.
@@ -69,9 +76,13 @@ Procedure:
                   "purpose": "…", "prep": "…",
                   "steps": {"before": [{"t": "…", "who": "agent"}], "during": [{"t": "…", "who": "pm"}], "after": [{"t": "…", "who": "agent"}]},
                   "initiative": {"id": "<initiative id>", "why": "…"},
-                  "links": [{"label": "Agenda", "url": "https://…"}]}],
+                  "links": [{"label": "Agenda", "url": "https://…"}],
+                  "refs": [{"label": "…", "source": "confluence", "link": "https://…", "lines": ["…"]}]}],
     "items": [{"title": "…", "why_now": "…", "source": "calendar|wispr|mail|slack|jira|confluence|github|drive|figma|website|derived", "initiative_id": "<id or omit>", "goal": "<goal id or omit>", "mention": false, "owner": "pm|agent", "effort": "minutes|an-hour|half-day|day", "confidence": 0.0, "evidence": "…",
-              "link": "https://…", "where": "#pm-team", "who": "…", "about": "…", "ask": "…", "reply_needed": true, "how": "…", "kind": "you|approve|auto"}],
+              "link": "https://…", "where": "#pm-team", "who": "…", "about": "…", "ask": "…", "reply_needed": true, "how": "…", "kind": "you|approve|auto",
+              "verb": "reply", "todo": "…", "due": "2026-10-07T15:00:00", "steps": [{"t": "…", "ref": 0}, "…"],
+              "thread": [{"who": "…", "at": "09:12", "text": "…"}], "refs": [{"label": "…", "source": "jira", "link": "https://…", "lines": ["…"]}],
+              "send": {"to": "…", "suggest": ["…"], "review": false}}],
     "digests": [{"source": "slack|jira|confluence", "name": "#pm-team", "link": "https://…", "decisions": ["…"], "open": ["…"], "goals": ["…"]}],
     "places": [{"source": "slack|jira|confluence", "name": "#pm-team", "link": "https://…"}],
     "memory": {"from": "YYYY-MM-DD", "to": "YYYY-MM-DD",

package/workspace/test-lists/today-act-prepare.json

@@ -0,0 +1,5 @@
+{
+  "_comment": "initiative today-act-prepare: the PM's marks on today's items and meetings (lib/day-marks.js, the mark and draft routes).",
+  "modules": ["day-marks.test.js"],
+  "seconds": { "day-marks.test.js": 8 }
+}