Skip to content
  1. Skills at a glance
  2. Report, review, get smoother over time
  3. /pstack:show-me-your-work

/pstack:show-me-your-work

Auto

In one lineHas the AI log one line per decision: what it did, why, the evidence, the result.

When to use
Long tasks, or tasks you'll leave running on their own.
Say it like this
/pstack:show-me-your-work Walk me through what you did last night

Just asking vs. using it

The same request, typed plainly or with /pstack:show-me-your-work. Here's what changes.

Just askingWith /pstack:show-me-your-work
Backs up its claims with paragraphs of prose.Logs one TSV row per decision, with a commit, PR or file:line as evidence.
Reviews its own work and calls it fine.Has a model from another family review the trail, then ends with an Attention section.
Trusts its own log as written.Checks every row against the run's transcript before handing back, adding rows for gaps.

What the document says

8 sections, about 1100 words, 5 min to read. Click a section to read it.

OpeningIt opens with one line: keep a single log.

  1. 1The formatOne TSV file, one row per decision, with six columns: ts, phase, decision, why, evidence, result. An example follows.
  2. 2Logging a rowAppends plain-worded rows with scripts/log.sh for forks, finished units and reverts. Each new run opens with a start row.
  3. 3Where it livesKeeps the log at decisions.tsv in the work dir, out of git, committed only when a reviewer needs the trail.
  4. 4RulesAppend only: a new row supersedes a wrong one. Evidence from committed scripts beats one-off checks.
  5. 5Audit the log against the transcriptBefore handing back, reads this run's transcript, checks each row and its evidence, and adds rows for gaps and errors.
  6. 6Cross-model review of the trailA helper on a different model family reviews the trail. The reply ends with an Attention section naming that model.
  7. 7Reviewing the trailTells the reader to read top to bottom, follow evidence pointers, and render the TSV with column in a terminal.
  8. 8Composing this skillOther skills point to this one for their audit trail and let it own the format.

See the wording worth borrowingRead the full text

Wording worth borrowing

A few sentences from the original: what each one makes the AI do, why it works, and how you can borrow it.

  1. Ask where the proof is

    Evidence is a pointer, not prose.

    Makes the AI
    Each row comes with a commit, file or screenshot you can open, not a paragraph saying "I checked".
    Why it works
    Fixing the shape of evidence leaves the AI no room to swap proof for nice prose.
    Borrow it
    When you ask for conclusions, add: "Put the file name and line number after each one; no need to explain."
  2. Pair each ban with an alternative

    A wrong call gets a new row that supersedes it.

    Makes the AI
    When an earlier row was wrong, it adds a correction below and leaves the old row, so you can see it changed its mind.
    Why it works
    A bare "don't edit history" leaves the AI stuck or tempted to cheat; naming what to do instead gives it a way forward.
    Borrow it
    When you forbid something, give the alternative: "Don't delete old entries; mark outdated ones with a new line 'void, see #5'."
  3. Pack rules into short contrasts

    Correct the log, not the story.

    Makes the AI
    When the check finds a mismatch, it adds a row with what really happened instead of polishing the story.
    Why it works
    "Fix this, not that" is short enough to remember, so the AI keeps it even at the end of a long run.
    Borrow it
    Write your most important rule as a short contrast: "Fix the bug, not the error message."
  4. Say "none" is an answer

    "No flags" is a valid value.

    Makes the AI
    When the review finds nothing, it writes "No flags" instead of inventing a few to look useful.
    Why it works
    Once an empty answer counts, the AI has no reason to make up problems.
    Borrow it
    When you ask for problems, add: "If you find no risks, say 'none'; don't pad the list."

Full text

The SKILL.md the AI reads, as written.

descriptionKeep a reviewable decision trail for long-running or unattended work: a TSV log with one row per decision (what, why, evidence, result). Local by default; commit it when a reviewer needs the trail to trust the result. Use for /pstack:show-me-your-work, autonomous or multi-phase runs, or work a human reviews after stepping away.

Show me your work

Keep one canonical log.

The format

A single TSV file, one row per decision. Cells stay single-line. Evidence is a pointer, not prose.

Copy references/decision-log-template.tsv (the header row) to start a clean log. Columns:

  • ts. ISO8601 timestamp.
  • phase. The phase or workstream.
  • decision. What was chosen or done, one line.
  • why. The reason in plain words. If a principle drove it, say it plainly, not as a jargon tag.
  • evidence. A link or path that proves it: commit SHA, PR number, file:line, or an artifact, trace, or screenshot path. Never a paragraph.
  • result. The outcome or predicate state: tests green, reverted, pixel-diff 0, INCONCLUSIVE, open.

An example, plain-spoken so a reviewer reads it at a glance.

ts	phase	decision	why	evidence	result
2026-05-24T09:02:00Z	frame	counted the work first, about 100 components and roughly 75 hours	wanted to know the size before starting a long run	commit 3a9f1c2	found 5 things to sort out before starting
2026-05-24T09:40:00Z	harness	took screenshots of the old version before changing anything	so we can compare old against new and catch any visual change	scripts/snapshot.sh, baseline/	saved 120 reference screenshots
2026-05-24T11:15:00Z	widget	moved the widget styles over without changing how it looks	keep the change small and the result identical	commit 7c21e0a, pixel-diff 0	looks identical, tests pass
2026-05-24T12:30:00Z	widget	threw out a helper's work because its screenshots were blank	checked the real files instead of trusting its summary	worktree reset	reverted, tightened the instructions for next time

Logging a row

Write each entry the way you'd tell a teammate what you did. Plain words, concrete actions, no AI speak or abstract jargon (the unslop skill applies to log text too).

Use the helper scripts/log.sh <logfile> <phase> <decision> <why> <evidence> <result>. It stamps ts, writes the header on first use, strips stray tabs/newlines, and prefixes any cell starting with =, +, -, or @ with a single quote. A bare printf appending a row works too, but mind those same bytes if cells come from generated or user-supplied text.

Log decision points and checkpoints, not every action: a fork chosen, a unit completed with its verification result, a pivot or revert with its trigger, a blocker surfaced, a gate fixed. For loop runs, one row per iteration. Skip the trivial and self-evident.

A run is one agent conversation, including its later turns and any summary of it. A pickup, a replacement agent, or a new chat starts a new run. When a run adds to a log that already has rows, its first row has phase start, and so does its first row after another run's start row. So a run that comes back to a log in a later turn first reads the log's last rows to see whether another run wrote since. A start row names the ts range of the rows before it that this run did not write, and its evidence names this run, such as its agent id. Use phase start for nothing else.

Where it lives

By default the log is a working artifact, not committed. Keep it at decisions.tsv in the work dir, or .audit/<task-slug>.tsv when several efforts run at once, and leave it out of git.

Commit it only when the work is ambitious enough that a reviewer needs the trail to trust the result.

Rules

  • Append-only. A wrong call gets a new row that supersedes it. Never edit or delete history.
  • Prefer evidence produced by committed scripts over hand-made one-offs (the encode-lessons-in-structure principle skill).

Audit the log against the transcript

At the end of the run, before handing back, check the log told the truth. Read this run's transcript from the active workspace's session directory: in Cursor, the agent-transcripts/ directory the system prompt names. In Claude Code, ~/.claude/projects/<slug>/ (the workspace path with every character that isn't a letter or digit turned into "-"). In Pi, ~/.pi/agent/sessions/--<slug>--/. In Codex, the ~/.codex/sessions/ file whose first line has payload.cwd equal to the workspace path. Don't glob across every project's directory. That reads unrelated private chats. Walk this run's rows against what actually happened. Each stretch of them begins at one of this run's start rows, or at the first row if this run created the log, and ends at the next start row of another run:

  • Check that every row maps to a real decision or action.
  • Check that each row's evidence resolves and shows what the row claims.
  • A fork, pivot, or abandoned approach that shaped the work but isn't logged is a gap. Add it.

Correct the log, not the story. The audit never edits or removes a row, even an invented one. When a row records neither a real decision nor a real action, or its claim or evidence is wrong, add a row that supersedes it with what actually happened and a pointer that resolves. This audit does not check rows outside this run's stretches. If this run's own work shows one of them is wrong, supersede it like any wrong call.

Cross-model review of the trail

Before handing back, spawn a subagent on a different model family from the one that did the work. Self-review is not a substitute. The subagent reads the audit trail and the run's transcript, then flags what the user should pay attention to. Not a redo of the work, a scan for what's suboptimal or risky. Pick each subagent's reasoning level and fast or standard speed for the job, per the Subagents section of the poteto-mode skill.

  • Decisions logged with weak or absent evidence.
  • Verification steps skipped or claimed without proof in the transcript.
  • Choices that look risky in hindsight (premature, scope-creeping, papering over a symptom).
  • Gaps the user would otherwise miss on a casual skim.

Every reply for a run that produced a trail ends with an "Attention" section. Lead with the reviewer's model on its own line (reviewed by <model>), then list each flag pointing to specific rows or moments. "No flags" is a valid value. The model name is not.

Reviewing the trail

Read top to bottom, follow the evidence pointers, spot-check. GitHub renders a committed TSV as a table. column -s$'\t' -t decisions.tsv renders it in a terminal.

Composing this skill

Other skills route their audit trail here instead of inventing one. Reference it by name and let it own the format. Don't restate the columns.

Original by Lauren Tan (poteto), MIT license. Checked against the pstack mirror at 97de40d. View on GitHub

An unofficial beginner's guide to pstack, not its official docs. pstack is made by poteto; the official repo is the source of truth for commands and behavior. The original guide source is on GitHub.

Hosted by airali. Adapted from Xiaoxiaodong's MIT-licensed guide.