csift SKILL.md skillA

csift from wdhwg001/csift is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add wdhwg001/csift/csift@git:20261002.d82d330

Pins git:20261002.d82d330 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
wdhwg001/csift
Path
skills/csift/SKILL.md
Versions recorded
2
Latest version
git:20261002.d82d330, committed Oct 2, 2026
First version recorded
Oct 1, 2026
Size
25,153 bytes
Repository stars
34
License
MIT

Reads, searches and counts Claude Code session transcripts (the .jsonl files under ~/.claude/projects, subagents too) with the csift command-line tool, and ships hooks that re-inject compacted turns or record pending questions. Use when asked what a past or running session said, did, edited, spent or is doing now; when about to grep, jq or script over a transcript jsonl (a type=="user" filter overcounts the person's messages), count how many messages the person typed, tool calls or tokens, or…

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# csift

csift reads Claude Code session transcripts (`~/.claude/projects/<encoded-cwd>/<session-uuid>.jsonl`, their `subagents/` transcripts and sidecar files) and answers with labelled records, counts and addresses you can re-run. The failure it prevents: a hand-written `grep`, `jq` or script over that format returns a plausible answer that is silently wrong, for the reasons law 1 lists.

Surface: csift 0.13.0 (must equal the output of `csift --version`, which prints `csift 0.13.0`). `csift <command> --help` is the authority on flag names and syntax, and when this skill and a help page disagree on those, the help page wins; on a behaviour this skill measured, the measured line wins, because a help page describes intent and a probe describes the binary. When a command you were confident about errors, compare `csift --version` with this line before anything else. Provenance labels: "measured" (with the command that ran on this surface), "docs" (a help page of this surface, a section of csift's SPEC.md at https://github.com/wdhwg001/csift/blob/v0.13.0/SPEC.md, or a Claude Code tool schema as a session of the named Claude Code version loaded it), UNTESTED (not run).
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add wdhwg001/csift/csift@sha256:f7dc06b0be7aeb50

The line at the top of this page pins the label git:20261002.d82d330, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_wnyhhivenydex6ok.svg)](https://markdownregistry.com/a/art_wnyhhivenydex6ok)

0 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 2 versions of this file, between Oct 1, 2026 and Oct 2, 2026, more than one a day.

versioncommittedcommitsizeaudit
git:20261002.d82d330 latest2026-10-02 d82d330 25,153 BA view · diff
git:20261001.ece0ec72026-10-01 ece0ec7 25,163 BA view

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (25153 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

wdhwg001/csift · 34 stars · license MIT · pushed 2026-10-04 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_wnyhhivenydex6ok
GET https://markdownregistry.com/api/v1/resolve?ref=wdhwg001/csift/csift
GET https://markdownregistry.com/api/v1/blob/f7dc06b0be7aeb50813bb119ea977d62b74e627629fcaeb0eac2e099e066bb42

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from wdhwg001/csift

AGENTS.md AGENTS.md file
wdhwg001/csift · AGENTS.md
git:20261002.2945d06 · audit B · 34 stars
CLAUDE.md CLAUDE.md file
wdhwg001/csift · CLAUDE.md
git:20260613.81a90be · audit B · 34 stars

Every file in wdhwg001/csift

Browse by kind, by grade A, or by owner.