| name | repost-history |
| description | Read the per-pair posted.jsonl history file and show recent entries. Use when the user asks "what did I repost from <pair>", "show <pair> history", or "tail repost history". |
| when_to_trigger | User wants to see what's been reposted from a given pair. |
Repost History
Tail the per-pair posted.jsonl and show recent entries.
Steps
-
Read the requested pair's posted.jsonl at ~/.repost-with-agent/pairs/<id>/posted.jsonl.
-
If file doesn't exist or is empty, tell the user "No posts yet for ." and stop.
-
Default tail length: 20 lines. Accept --limit N from the user.
-
Read with tail -<N> via Bash. Each line is a JSON object with fields:
ts (ISO-8601)sourceItemId (e.g. urn:li:activity:7000)canonicalSourceUrldestinationUrldestinationIddraftText (the text we posted, ≤500 chars summary if longer)
-
Format each entry compactly:
2026-04-30T22:53Z urn:li:activity:7000
Source: https://www.linkedin.com/feed/update/urn:li:activity:7000/
Posted: https://x.com/REEEthan_YT/status/123
Text: "Yesterday I shipped v3.0.0 of Repost-with-agent — instructions + JSON state, no more SDKs..."
Optional: include a learnings tail
If the user asks for --with-learnings (or similar) on top of the history
tail, also read ~/.repost-with-agent/pairs/<id>/learnings.md and print the
last 3 ## … entries under a === Recent learnings (last 3) === heading
below the post history. This is handy when the user is reviewing a sequence
of posts and wants the institutional context that informed them.
If learnings.md is empty or only contains the placeholder stub, print
(no learnings recorded yet). See skills/repost-learnings/SKILL.md for
the file format.
By default (no --with-learnings), this skill is a pure history tail and
doesn't touch learnings.md.
See also
skills/repost-pair-show/SKILL.md — full pair inspection (config + history
skills/repost-learnings/SKILL.md — format of the learnings.md file.
Telegram
Read-only. No Telegram on this skill.
