agentleFS
priyanshu-aryapriyanshu-arya/LinkedIn_Automation/.claude/skills/plan-week/SKILL.md

plan-week

Skill0 starsChanged 59 days agoMITView on GitHub

Use when the user asks to plan the week's LinkedIn content, decide what gets posted when, run the Content Strategist/Content Planner, or explicitly invokes /plan-week. Produces a 7-day calendar VIEW (Mon-Sun) but only fills the quality-gated slots that clear the bar (default ~3/week, Tue/Thu/Sat starting heuristic) per REQUIREMENTS.md §5 -- remaining days are shown explicitly empty with a stated reason, never padded. Each filled day gets a topic, format, a hook formula (from Content-Learnings/hook-formulas.md), a posting time (per §11, overridden by playbook.md evidence), and 2-3 comment-target suggestions. Does not draft post text (that's /write-draft).

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 4 are distinctive, and 9 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 9

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Default 3. This is a ceiling to aim toward, not a quota to force — see the hard rules below.

    line 29
  2. Very specific:

    Read all Idea Notes in Post-Ideas/ with status: candidate (never placeholder).

    line 35
  3. Very specific:

    Days: read Content-Learnings/playbook.md first.

    line 49
  4. Very specific:

    Fit check: prefer a canonical formula whose engagement_goals value matches the idea's likely engagement goal — the same precedence …

    line 85
  5. Very specific:

    This week's own picks: never assign the same formula_id to two different filled slots within the same plan produced by this run.

    line 95
  6. Very specific:

    If content-index.md doesn't exist yet or is too thin to judge from, this check is a no-op — don't block on it, same fallback posture as …

    line 102
  7. Distinctive:

    No genuine fit: if nothing in the canonical set fits this idea's category/content_type (a goal mismatch alone isn't disqualifying if the …

    line 105
  8. Very specific:

    Override: if Content-Learnings/playbook.md has a Best-Performing Patterns rule with real evidence (≥3 published posts, post ids attached) …

    line 124
  9. Distinctive:

    Neither source yields anything: leave the field explicitly empty in the report with a one-line reason (e.g.

    line 146
  10. Very specific:

    If fewer than target_count slots ended up filled because the pool genuinely didn't support more distinct, valuable angles, say that plainly …

    line 166
  11. Very specific:

    Never assign more than target_count (default 3) posts in one week, and never assign to Sunday unless the user explicitly asks for it this …

    line 173
  12. Distinctive:

    Never draft post text here — ideas only get a day, not content.

    line 179
  13. Very specific:

    Never pick a status: placeholder idea.

    line 180

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Plan Week (Content Strategist / Content Planner — Phase 4a, revised Phase 13, reworked Phase 21)

Assigns ranked candidate ideas to specific days under the flexible weekly cadence (REQUIREMENTS.md §5), then — for each slot that actually gets filled — attaches a suggested hook formula, a posting time, and 2-3 comment targets (REQUIREMENTS.md §32), and renders the whole thing as a full Mon-Sun calendar view with unfilled days shown explicitly empty rather than hidden. Selection and scheduling-day decisions only — does not write post text (that's /write-draft) and does not touch Buffer or actual scheduling (that's a later phase). §32 is additive to this file's original §5 logic: it changes nothing about which ideas get picked or how many slots get filled, only what gets attached to a slot once it's filled and how the result is displayed.

Arguments

  • week (optional) — an ISO week (e.g. 2026-W38) or a Monday date to plan for. If omitted, default to the next full week from today's actual date (if today is on or before Saturday of a week that hasn't started posting yet, that week; otherwise the following week). Never plan into a week that already has target_count slots filled with status: selected or later.
  • target_count (optional) — how many posts to aim for this week. Default 3. This is a ceiling to aim toward, not a quota to force — see the hard rules below.

Process

1. Load the candidate pool

Read all Idea Notes in Post-Ideas/ with status: candidate (never placeholder). Note each one's category, content_type, format, rank_score.

2. Check recent history for fatigue

First scan Content-Learnings/content-index.md (fast, compact) for category/format/topic repetition in the prior ~2 weeks. For anything it flags as close, confirm against the actual note (Drafts/, Scheduled/, Published-Posts/). Avoid assigning two similar posts back-to-back across week boundaries, not just within the week being planned. If content-index.md doesn't exist yet or is still thin, fall back to reading the folders directly, as before.

3. Pick the days, then apply the content-type variety rule

Days: read Content-Learnings/playbook.md first. If it has a Best-Performing Patterns rule about which specific days perform better, with real evidence (post ids, ≥3 published posts) attached, use those days instead of the default below. Otherwise fall back to the starting heuristic (REQUIREMENTS.md §5): Tue, Thu, Sat, trimmed to however many slots target_count actually calls for (e.g. target_count: 2 → Tue/Thu). Sunday is never used unless explicitly requested this invocation.

Content-type variety: unlike the old fixed Mon–Fri table, a specific content type is no longer bound to a specific day. Instead, across however many slots this run fills, require at least 2 distinct category/ content_type values among them (checked against the Playbook's evidenced patterns first, this variety rule otherwise) — don't let 3 slots accidentally all be the same content type.

4. Match ideas to slots

For each day slot, pick the unassigned candidate idea whose category/ content_type best fits, breaking ties by rank_score, while keeping the variety rule from step 3 satisfied across the slots filled so far. Do not force a mismatched or duplicate idea into a slot just to fill it. If no good fit exists for a slot, leave it unfilled.

5. Commit the assignment

For each idea assigned to a slot:

  • status: candidate → selected
  • target_week: the ISO week being planned
  • target_date: the specific date

6. Assign a hook formula (filled slots only, REQUIREMENTS.md §32)

For each slot filled by step 5 — never for a slot left empty — read Content-Learnings/hook-formulas.md fresh (not from memory; it grows over time via /extract-hook and human promotions). Filter to status: canonical rows only — a proposed row is not trusted for use here any more than it is in /write-draft.

  • Fit check: prefer a canonical formula whose engagement_goals value matches the idea's likely engagement goal — the same precedence /write-draft step 3 uses, but non-binding here: the idea's own goal field if it has one and is set, else a Content-Learnings/playbook.md- evidenced pattern for the idea's category (real post ids attached as evidence), else default comments — and whose mechanic genuinely fits the idea's category/content_type. A formula with no engagement_goals value is general-purpose and eligible regardless of goal.
  • Anti-repetition — two checks, both required:
    1. This week's own picks: never assign the same formula_id to two different filled slots within the same plan produced by this run.
    2. Recent history: scan Content-Learnings/content-index.md's hook_gist column (§15/§25.4) for the last few weeks of published/scheduled posts and avoid a formula whose mechanic was just used — judge this from the hook_gist text and, where a row's linked note is easy to open, that note's own hook_formula field. If content-index.md doesn't exist yet or is too thin to judge from, this check is a no-op — don't block on it, same fallback posture as step 2's fatigue check.
  • No genuine fit: if nothing in the canonical set fits this idea's category/content_type (a goal mismatch alone isn't disqualifying if the mechanic still fits — only a real mechanic mismatch is), leave that slot's hook-formula cell blank in the report and say why in one line — never force a bad fit just to populate the column.
  • This assigns a formula suggestion, not the finished hook line — /write-draft re-reads hook-formulas.md itself at drafting time and may pick differently if circumstances have changed by then (a new proposed formula got promoted, playbook evidence shifted, etc.). Nothing here binds /write-draft's own step 5.

7. Assign a posting time (filled slots only, REQUIREMENTS.md §32)

For each filled slot, attach the time exactly as REQUIREMENTS.md §11 already specifies — pure surfacing of existing logic at planning time instead of leaving it until /schedule-approved; this step adds no new rule of its own:

  • Default: the day's row from §11's table for whichever day step 3 assigned this slot (e.g. Tuesday → 16:00, Thursday → 17:00, Saturday → 09:00).
  • Override: if Content-Learnings/playbook.md has a Best-Performing Patterns rule with real evidence (≥3 published posts, post ids attached) for a better day+time+content-type combination covering this slot, use that time instead — the same override step 3 already applies when choosing which days to use; this step applies it again to the time attached to whichever day the slot ends up on.

8. Assign comment targets (filled slots only, REQUIREMENTS.md §32)

For each filled slot, attach 2-3 specific external accounts/posts to proactively comment on that day, in this priority order:

  1. Content-Learnings/comment-targets.md, if it exists and its ## Targets table has rows — a user-maintained list (new in Phase 21, never system-populated). Prefer entries relevant to the slot's topic/category; if several are equally relevant, prefer ones not already used earlier in this same week's plan.
  2. Fallback, only if (1) is missing, absent, or too thin to fill the slot: accounts/authors/publications that came up as sources or citations in the research actually backing that slot's idea — i.e., a named author or account identifiable in the sources[] of the Research Notes the idea itself cites from that week's /research-topic runs. Never a generic outlet name with no identifiable account, and never a source from an unrelated topic.
  3. Neither source yields anything: leave the field explicitly empty in the report with a one-line reason (e.g. "no comment-targets.md entries and no research-cited accounts this week") — never invent a plausible-sounding account or post to fill the cell. This is exactly as unacceptable as forcing a hook-formula fit or padding the day count.

9. Report

Render a Mon-Sun calendar table, all seven days, not just the filled ones: date | day | status | topic | category | format | hook formula | posting time | comment targets | rank_score.

  • Filled rows (status: filled): every column populated from steps 4-8.
  • Empty rows (status: empty): the reason already produced for that gap by step 4 goes directly in the row (e.g. "no Career-category ideas in the pool — run /research-topic Career then /generate-ideas before this slot can be filled") — shown in the row itself, not buried in separate prose below the table. hook formula and comment targets are blank on every empty row by construction (steps 6 and 8 only ever touch filled slots) — never populate them for a day that didn't clear the bar.
  • If fewer than target_count slots ended up filled because the pool genuinely didn't support more distinct, valuable angles, say that plainly in the table and in a closing line — that is the correct outcome per §5, not a shortfall to apologize for.

Hard rules

  • Never assign more than target_count (default 3) posts in one week, and never assign to Sunday unless the user explicitly asks for it this invocation. Saturday is in scope by default.
  • Never force the count. Filling fewer slots than target_count because nothing else cleared the bar is the correct, expected outcome — never pick a weak or duplicate idea just to reach the target.
  • Never draft post text here — ideas only get a day, not content.
  • Never pick a status: placeholder idea.
  • Hook formulas and comment targets are attached only to filled slots (REQUIREMENTS.md §32). Never invent either for an empty day, and never let hook-formula or comment-target scarcity be a reason to fill a slot that otherwise wouldn't clear the quality bar. This rule carries exactly the same weight as the count-padding prohibition above — it is not a softer, secondary concern.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.