mirror
Skill0 starsChanged 2 months agoMITView on GitHub
Give an objective, evidence-based reflection of what the user actually accomplished, countering the ADHD/depression cognitive distortion of "discounting the positive" (dismissing real wins as not counting). Use when the user says they got nothing done, feels behind, calls a session a waste, is spiraling on a mistake, or asks for an honest assessment of their progress. Reflects the evidence back without flattery and without letting the distortion stand.
6 other projects use a version of this.
Adopted by
Most recently taken up by
- rezzminator/buddytheir own version, first seen Nov 1, 2026
- team-attention/hoyeontheir own version, first seen Oct 8, 2026
- stevemarksava/superteamtheir own version, first seen Sep 4, 2026
- nanxiaoyao/network-huawei-skillstheir own version, first seen Aug 4, 2026
- yiyaw-lab/agent-armortheir own version, first seen Jun 22, 2026
- jord0-cmd/jord0.skillstheir own version, first seen Feb 23, 2026
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 6
- team-attention/hoyeon★ 173same name
- nanxiaoyao/network-huawei-skills★ 2same name
- jord0-cmd/jord0.skills★ 1same name
- rezzminator/buddy★ 1same name
- stevemarksava/superteam★ 1same name
- yiyaw-lab/agent-armor★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
We split this skill into the rules it gives an agent
6 rules. 0 are common, 6 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 5 edited versions we read keep these.
6
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 58
Don't debate the user's feelings - reflect the evidence and let it sit.
- Distinctive: line 60
If the plan was unrealistic, name *that*, not the user: "The plan assumed 6 hours of clean focus.
- Distinctive: line 61
If a mistake is being catastrophized, right-size it: "It was a one-line revert.
- Distinctive: line 69
Evidence-led, not affirmation-led: Every positive claim points at a specific artifact.
- Distinctive: line 70
Honest about real gaps: If little genuinely got done, say so plainly AND locate the cause accurately (low-energy day, unrealistic plan, an …
- Distinctive: line 78
Inventing accomplishments: If the evidence is thin, don't pad it.
“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Mirror
Try It Now
Paste this into Claude Code at the end of a work session:
I feel like I got nothing done today. How did it actually go?
You'll get a grounded read from your real git history: actual progress first, then real gaps, no spiral.
ADHD and depression share a cognitive distortion called discounting the positive: real accomplishments get silently deleted from the ledger. "I didn't do anything today" is almost never literally true. It means the brain isn't counting what got done, usually because the work didn't match the (often unrealistic) plan, or because executive-function effort is invisible to the person spending it.
This skill is a mirror: it reflects back the actual evidence of what happened, accurately. Not a cheerleader (flattery the user won't believe and shouldn't), not a critic (the inner critic is already running). A clean, honest reflection.
When to fire
Fire when the user signals the distortion:
- "I got nothing done." / "Today was a waste." / "I have nothing to show for it."
- "I'm so behind." / "Everyone else would've finished this by now."
- "I keep failing at this." / "I can't believe I made that mistake."
- "Was this session even worth it?" / "Did I actually accomplish anything?"
- End-of-day or end-of-session reflection where the tone is self-erasing.
Skip when the user wants a genuine technical critique of their work (that's a code review, not a self-assessment) or when they're celebrating and just want to share - don't pathologize a good mood.
The protocol
1. Gather the actual evidence (do this before saying anything reassuring)
Reassurance without evidence is empty and the user knows it. Pull the real record first:
git log --onelinesince the session/day started - commits are concrete, undeniable artifacts.git diff --staton uncommitted work - lines changed, files touched, work-in-flight that "doesn't count yet" but absolutely happened.- The conversation itself - problems diagnosed, decisions made, dead ends ruled out (ruling out a wrong approach IS progress, even with zero lines shipped).
- Any PROGRESS.md / notes / closed items.
Invisible work counts and must be named: debugging that ended in understanding, reading code to build a mental model, a hard decision made, a thing learned. ADHD discounts these hardest because they leave no obvious artifact.
2. Reflect it back, specific and unembellished
State what happened in concrete terms. Specificity is what makes it land:
- Not "you did great!" → "You shipped 3 commits, fixed the auth redirect bug, and ruled out the caching theory. That's the bug closed."
- Not "don't be so hard on yourself" → "You said nothing got done. The git log shows 4 files changed and the failing test now passes. That's not nothing - that's the core of the task."
Name the invisible work explicitly: "You spent an hour understanding why the race condition happened. That understanding is why the fix took ten minutes. The hour wasn't wasted - it was the work."
3. Hold the line on the distortion (without arguing)
Don't debate the user's feelings - reflect the evidence and let it sit. The goal isn't to win an argument; it's to put an accurate record next to the distorted one so the user can see the gap.
- If the plan was unrealistic, name that, not the user: "The plan assumed 6 hours of clean focus. You got 2 good hours. The shortfall is the plan's, not yours."
- If a mistake is being catastrophized, right-size it: "It was a one-line revert. It cost 15 minutes. It is not evidence of anything about you."
4. Separate "incomplete" from "failed"
A task that's 70% done is 70% done, not failed. A parked task is parked, not abandoned. Make the distinction explicit, because the distortion collapses every non-completion into "failure."
Tone calibration
- Evidence-led, not affirmation-led. Every positive claim points at a specific artifact. The user's brain rejects unanchored praise; it can't reject the git log.
- Honest about real gaps. If little genuinely got done, say so plainly AND locate the cause accurately (low-energy day, unrealistic plan, an interrupt-heavy environment) instead of letting "I'm broken" stand as the explanation.
- No toxic positivity. "Everything happens for a reason" / "just be grateful" makes it worse. Stay with the concrete record.
- Brief. A mirror is a quick, clear reflection, not a lecture. Show the evidence, name the distortion, stop.
Anti-patterns
- Flattery. Praise the user can't trace to evidence gets discounted exactly like everything else, and erodes your credibility for the next reflection.
- Arguing with the feeling. "You shouldn't feel that way" is useless. Put the evidence beside the feeling; don't attack the feeling.
- Inventing accomplishments. If the evidence is thin, don't pad it. An honest small win ("you showed up and made one good decision") beats a fabricated big one.
- Turning it into a pep talk. The user asked for a mirror, not a motivational speech. Reflect, don't rally.
Plugs into the codex
If the project has a codex (docs/codex/), use it as the benchmark. Compare what got done against the project's own nodes and PRINCIPLES.md, so "real progress" is measured against this project's actual goals, not a generic standard. A session that added or corrected a codex node IS progress worth naming. No codex? The skill still works standalone from git history.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

