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

plan-week-x

Skill0 starsChanged 59 days agoMITView on GitHub

Use when the user asks to plan the week's X (Twitter) content, decide what gets posted when on X, or explicitly invokes /plan-week-x. Assigns candidate ideas tagged for the x platform from Post-Ideas/ to specific days under X's own cadence (default ~5/week, mix of singles and threads), without drafting any post text.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 3 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Default 5. A ceiling to aim toward, not a quota to force.

    line 19
  2. Very specific:

    Read all Idea Notes in Post-Ideas/ with status: candidate and x in their platforms list.

    line 25
  3. Very specific:

    Days: read Content-Learnings/playbook-x.md first — if it has a day-performance rule with real evidence (≥3 published X posts) attached, use …

    line 39
  4. Very specific:

    Never assign more than target_count (default 5) X posts in one week.

    line 92
  5. Distinctive:

    Never force the count — fewer filled slots than target is the correct outcome when the pool doesn't support more.

    line 93
  6. Distinctive:

    Never draft post text here — ideas only get a day.

    line 95
  7. Very specific:

    Never pick a status: placeholder idea, or an idea without x in its platforms list.

    line 96

≈ “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 X (Content Strategist — X pipeline)

X's sibling to /plan-week. Same algorithm, same shared candidate pool (Post-Ideas/) — only the cadence, day-heuristic, and platform filter differ. Selection and scheduling-day decisions only; does not write post text (that's /write-draft-x) and does not touch Buffer.

Arguments

  • week (optional) — an ISO week (e.g. 2026-W38) or a Monday date. If omitted, default to the next full week from today, same rule as /plan-week.
  • target_count (optional) — how many X posts to aim for this week. Default 5. A ceiling to aim toward, not a quota to force.

Process

1. Load the candidate pool, filtered to this platform

Read all Idea Notes in Post-Ideas/ with status: candidate and x in their platforms list. An idea with platforms: [linkedin, x] is eligible here too — it just gets its own independent draft via /write-draft-x, not a copy of the LinkedIn version.

2. Check recent history for fatigue

Scan Content-Learnings/content-index.md, filtered to platform: x rows, for topic/format repetition in the prior ~2 weeks. Fall back to reading Drafts//Scheduled//Published-Posts/ directly (filtered to platform: x) if the index is thin.

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

Days: read Content-Learnings/playbook-x.md first — if it has a day-performance rule with real evidence (≥3 published X posts) attached, use those days. Otherwise fall back to a starting heuristic sourced from current (2026) X/Twitter best-posting-time research — look this up live via WebSearch when this skill first runs rather than reusing LinkedIn's Tue/Thu/Sat/IST table, since X's audience behavior and this account's own mix of singles/threads don't necessarily match LinkedIn's pattern. Record whatever heuristic gets sourced back into this file's own notes so future runs don't re-research it every time, and say plainly that it's an unvalidated starting point until playbook-x.md has real evidence.

Sourced 2026-09-14 (aggregated from Buffer's and SocialPilot's 2026 studies, 700k+ and 1M+ post datasets respectively): Tuesday 9:00 AM is the single strongest slot; Wednesday 9-10 AM is a close second; Tue-Thu overall outperforms other weekdays; peak windows otherwise run 8-11 AM and 12-6 PM on weekdays. Starting default until playbook-x.md has real evidence: Tue/Wed/Thu 09:00 local. This does not need re-researching on every run — reuse this note, refresh it only if it's been a long time or the account's own data starts contradicting it.

Content-type & format variety: across however many slots this run fills, require at least 2 distinct category/content_type values, same rule as /plan-week. Additionally, don't let every slot default to the same single-vs-thread shape — /write-draft-x makes the final call per idea, but flag here if the pool looks like it would produce, say, 5 threads in a row with no singles.

4. Match ideas to slots

Same tie-breaking as /plan-week: best category/content_type fit, rank_score as tiebreaker, variety rule respected. Never force a mismatched or duplicate idea into a slot.

5. Commit the assignment

For each idea assigned to an X slot, append an entry to the idea's platform_schedule list: {platform: x, week: <ISO week>, date: <date>}. Never touch target_week/target_date — those belong to /plan-week's LinkedIn assignment exclusively, even for an idea that isn't going to LinkedIn at all (they just stay blank in that case). If status is still candidate (no platform has claimed it yet), advance it to selected; if another platform already advanced it, leave status as-is — status reflects "has at least one platform claimed this," not "claimed by every platform in its platforms list."

6. Report

A table: date | day | idea (topic) | category | format (single/thread judgment call, finalized at write-time) | rank_score — with unfilled slots called out plainly, same honesty rule as /plan-week.

Hard rules

  • Never assign more than target_count (default 5) X posts in one week.
  • Never force the count — fewer filled slots than target is the correct outcome when the pool doesn't support more.
  • Never draft post text here — ideas only get a day.
  • Never pick a status: placeholder idea, or an idea without x in its platforms list.

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.