bereach-ai/linkedin-prospecting-skills / skills/linkedin-outreach-method
linkedin-outreach-method
VerifiedRuns the whole LinkedIn outreach loop: find, sort by persona, voice, one template per persona, send, measure acceptance and replies at 7 days. Start here for any LinkedIn outreach.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 10.0 KB |
| references/advanced.md | markdown | 1.9 KB |
| references/artifacts.md | markdown | 19.9 KB |
| references/conversation-review.md | markdown | 4.7 KB |
| references/frameworks.md | markdown | 21.0 KB |
| references/icp.md | markdown | 20.0 KB |
| references/measuring.md | markdown | 10.2 KB |
| references/search-doors.md | markdown | 3.7 KB |
references/search-doors.md
View on GitHub# Search doors
Who is already in the account comes first; a search only fills a persona short of people. For that
persona the plan is the base; a card's other fields pick a second door.
## Existing contacts first (free)
1. The count, before the question round: `bereach_contacts` with `engaged: false`, `outreachStatus: "none"`,
`lifecycleStage: "contact"`, `doNotContact: false`, `limit: 1`, and its `total`. Keep `engaged: false`: without it the call returns only people already touched, and
nothing fails.
2. The split, after the answers, never a download of every row. The filters that exist: the four above,
`hasProfileData: true`, `campaignId` (one list), `tag`, `sortBy` (`profileUpdatedAt_desc` first),
`limit` up to 100, then the `nextPage` it answers with; there is no title, headline or place filter. A list whose people fit
one persona is paged alone with its `campaignId`; otherwise pages of 100, each from the last page's `nextPage`. Each row goes to a card by its
`headline`, `currentRole`, `company` and place against Found by and the place answered. Stop as soon as
every persona has a few hundred (enough for the first weeks; close enough moves on), at most 10 pages (1,000 rows);
the rest waits for a later batch, and the page counts only the rows read. A size or place the row does
not show is not looked up: the row stays in. No profile data: the same count call with `hasProfileData:
false`, its `total` is the amber `pas de profil : N`, never read one by one.
3. Before the invitations: the same call with `engaged: true` returns people touched in a way their status
does not show yet: take their ids out, and anyone in `bereach_invitations` with `direction: "sent"`.
4. Per list: the list ids from each row's `campaigns`, then `bereach_contacts` with `campaignId`,
`outreachStatus: "none"`, `limit: 1`, and read `total`.
5. On the Plan and personas yes (no search: as the page is shown): one `bereach_contact_update` per persona, up to 500 `contactIds`,
`tags: { add: ["persona:<short-name>"] }`. A row already tagged is skipped, so a split resumes.
A persona short of people: the doors below.
## The doors
| Card field or ask | Door | Saves to the list |
|---|---|---|
| Found by (titles, seniority) inside the ICP: always first for a persona short of people | `bereach_plan_searches`, then `bereach_run_searches` | yes |
| Why now or Signal is a company event (a raise, a new executive) | `bereach_find_signals`: events first, read them, then resolve the chosen companies with the card's role | yes, at resolve |
| Why now or Signal is a post on a topic | `bereach_find_posts` with `since`, `authorTitle`, `authorLocation`; then `bereach_collect_engagers` on the two or three best posts | authors yes unless `save: "skip"`; engagers always |
| Proof that lands names past clients | `bereach_find_people` with `likeCompanies` (companies like them, never them) | yes |
| Recently funded | `bereach_find_companies` with `filters.fundedWithinMonths`, then `bereach_find_people` with those companies and `perCompany` | companies no, people yes |
| Named accounts | `bereach_find_employees`, or `bereach_find_people` with the company and a role | yes |
| The list is built, a dated reason per person | `bereach_add_signals` (not a search), run for the whole list without a yes | writes each person's reason on the row |
- Each door's people are one row on Plan and personas: persona, why them in plain words ("levée
récente", "ont réagi à un post sur ..."), found. Never a tool name.
- A signal lead is tagged `signal:<name>` and invited first; its event or post is its line.
- Hiring and job ads are a weak signal: say so before running one. Events are reasons, never a count.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |