Harmix/open-stanley / open-stanley/skills/stanley
stanley
VerifiedYour head of content. Use this whenever the user wants help with LinkedIn or X (Twitter) — writing or scheduling a post, finding things to comment on, turning a meeting, voice note, or Slack thread into a post, reviewing why a post did or didn't work, planning the week's content, or asking "what should I post". Also use it for anything that mentions Stanley, ghostwriting, personal brand, founder-led content, building in public, or growing followers, even if the user doesn't say "post". Start here; it routes to the other open-stanley skills and holds the operating rules they all share.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 14.2 KB |
| references/capabilities.md | markdown | 3.3 KB |
| references/platform-access.md | markdown | 4.0 KB |
| references/recommended-mcps.md | markdown | 5.5 KB |
| references/recommended.mcp.json | config | 2.0 KB |
| references/research-linkedin-2026.md | markdown | 6.7 KB |
| references/scheduling.md | markdown | 8.6 KB |
| references/vault-backends.md | markdown | 6.8 KB |
| references/verify-live.md | markdown | 2.2 KB |
| references/why-these-rules.md | markdown | 2.1 KB |
| scripts/stanley-auditexecutable | script | 24.6 KB |
| scripts/stanley-briefexecutable | script | 6.4 KB |
| scripts/stanley-chartexecutable | script | 14.5 KB |
| scripts/stanley-previewexecutable | script | 5.8 KB |
| scripts/stanley-statsexecutable | script | 14.9 KB |
| scripts/stanley-trace | other | 6.3 KB |
| scripts/stanley-vaultexecutable | script | 35.1 KB |
references/platform-access.md
View on GitHub# Platform access: official API/MCP before the browser
The browser is the fallback, not the default. Before automating anything on a platform, ask which of two things you are doing, because the answer is different for each:
- **Write** — post, comment, like, reply *as the user, on the user's own behalf*.
- **Read** — the feed, search, other people's posts, analytics.
Then check, in this order: (1) an MCP the user already has connected that covers it (`post` capability in `capabilities.md`); (2) an official first-party API or hosted MCP the user could connect in ten minutes; (3) the browser. Record the answer per platform in `my-human.md` as `post_path: <platform>=<mcp|api|browser>` so later runs do not re-derive it.
## Why writing through the browser is the risky half
Reading through a logged-in browser looks like a person scrolling. Posting through one is the part that looks like automation to a platform's own detectors: the typing is machine-fast, the timing is exact, and it arrives from a session that never idles. Both platforms below publish rules against third-party automation of the account, and both offer an official write path to your *own* profile that is free or nearly free. So when a write has an official path, use it.
Data point, not a proven cause: this plugin posted a four-tweet X thread through the browser on 12 September 2026; the account was suspended the following day. It may have been unrelated. It is still the reason the order above is what it is.
## LinkedIn (September 2026)
| | Path | Notes |
|---|---|---|
| Write | **Official, self-serve.** A LinkedIn developer app with the `w_member_social` scope posts, comments and likes on the authenticated member's own behalf. No partner approval, no fee. | The only open scopes are OpenID Connect sign-in and `w_member_social`. Setting it up is a one-time thing the user does; offer it once at onboarding and record the answer. |
| Read | **Browser.** | Feed reading, member/content search, and analytics on anyone (including the user's own creator analytics) all sit behind the Marketing/partner products, which need an application and approval. There is no self-serve read surface, so scout, recap and onboarding read through Claude in Chrome. This has worked without friction. |
## X (September 2026)
| | Path | Notes |
|---|---|---|
| Write | **Official hosted MCP**, `https://api.x.com/mcp` (first-party, shipped 30 June 2026). Posts, replies, bookmarks, and draft+publish Articles. Writes need OAuth 2.0. | The MCP layer is free; calls bill against the user's X API plan. The free tier covers roughly 500 posts/month of writes, which is far more than this plugin will ever use. Prefer it over the browser for every X write. |
| Read | **Browser**, unless the user pays. | The free X API tier includes no reads, so full-archive search, timelines and analytics through the MCP require Basic (~$200/mo) or pay-per-use credits. Not worth it for one person's content loop: read X in the browser, write through the MCP. |
X's MCP also exposes trends and news, which the scout can use for "what is happening today" *if* the user is on a paid tier. Check the tier before promising it.
## Any other platform
Same procedure, no guessing: search the user's connected MCPs for the platform by name, then look for a first-party API or hosted MCP, then fall back to the browser. Say in one line which path you used and why. Never scrape a platform that offers a self-serve write API for the thing you are about to do.
## When the browser is the only path
- Human pacing: one action at a time, a real pause between them, never a burst of posts or comments.
- Never post the same text to two accounts, and never run the same action on a loop.
- Keep to the `scout` skill's cap (25 posts opened per run) and the strategy's comment budget.
- One session, the user's own logged-in profile. No second account, no credential sharing.
- If a platform shows a captcha, a rate-limit notice, or a "confirm you are human" interstitial, stop, report it, and do not retry. That is the platform saying no.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |