Session Doubt
End-of-session doubt enumeration before claiming substantive work done. Surfaces (1) what the agent is least confident about — then root-causes each item — and (2) the biggest thing the user may not realize about the situation. A two-question reflective close-out, distinct from test-running verification. Triggers "/session-doubt", "doubt check", "what are you least confident about", "what am I mi…
Install
$ agentstack add skill-ao92265-claude-code-playbook-session-doubt ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Links to your public security report.[](https://agentstack.voostack.com/security/report/skill-ao92265-claude-code-playbook-session-doubt) Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
session-doubt
Reflective close-out for substantive work. Two questions, asked in order. Source: end-of-session technique (Sam Altman's Q1 + a Claude-suggested Q2) — "1 in 4 times it surfaces a huge deal."
Run AFTER the work is otherwise complete (and ideally after /done / verification-before-completion has run). This is reflection, not test execution — the two are complementary, not substitutes.
Q1 — Least confident
Ask yourself, honestly and without flattering the work:
> What am I least confident about right now? Enumerate everything.
- List every item, not a curated top-3. Aim for 5–8. Include things that feel minor.
- For each: one line stating the doubt + why.
- Then root-cause the real ones: pick the items where being wrong would actually matter and investigate each exhaustively (read the code, run the check, trace the path) until you understand the root cause — do not stop at restating the doubt.
- Be specific. "Error handling might be incomplete" is useless. "If the upload retries after a 413, we double-charge because the idempotency key is regenerated per-attempt — unverified" is useful.
Q2 — Biggest blind spot
Then ask:
> What is the biggest thing the user does not realize about this situation right now?
- One thing. The single highest-leverage reframe, risk, or connection the user is most likely missing.
- It should change a decision, not just inform. Prefer the insight that contradicts or reframes something the user (or you) already assumed.
Output
**Q1 — least confident (root-caused):**
1. — →
2. ...
**Q2 — biggest blind spot:**
Keep it tight. No reassurance padding. If genuinely nothing is uncertain, say so plainly — do not manufacture doubt to fill the list.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: ao92265
- Source: ao92265/claude-code-playbook
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.