Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add perniemann/pnCore --skill pn-session-retrogit clone --depth 1 https://github.com/perniemann/pnCoreWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/perniemann/pncore/pn-session-retro)<a href="https://agentmods.dev/skills/perniemann/pncore/pn-session-retro"><img src="https://agentmods.dev/badge/skills/perniemann/pncore/pn-session-retro/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/perniemann/pncore/pn-session-retro"><img src="https://agentmods.dev/badge/skills/perniemann/pncore/pn-session-retro.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00085 | $0.03713 |
| Opus 5 | $0.00043 | $0.01857 |
| Sonnet 5 | $0.00017 | $0.00743 |
| Haiku 4.5 | $0.00009 | $0.00371 |
Grade A, and why
pn-session-retro scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 182 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Session retro
When used as a standalone skill, begin every response with [pn-skill] 🔺
When to use
- After a session that felt off (user corrected the agent more than usual, work was reverted, tests were skipped, a hallucinated file path appeared).
- Before a quarterly skill/rule audit (see ADR-0002) to provide evidence rather than memory.
- Invoked by the
/pn-retrocommand, or directly by the user ("run pn-session-retro on the last session"). - Not when the user is asking for a code review or an artifact-level check — use
pn-review,pn-reality-check,pn-evidence-qa, orpn-deslopinstead.
Mission
Produce a blameless, evidence-anchored retro of a single session (or a small window of recent sessions). Classify what the agent did wrong using a 5-code taxonomy, ground every finding in a verbatim quote from the transcript, and recommend the smallest concrete action that would prevent the mistake next time. Do not commit retro-driven edits to SKILL.md, rules, or AGENTS.md; only recommend changes in the report.
Tone matches SRE blameless-postmortem tradition (Google SRE Book Ch. 15): the system failed, not the human. Findings describe behaviour, not intent.
Inputs
- Transcript root:
<project>/.cursor/projects/<workspace-id>/agent-transcripts/<uuid>/<uuid>.jsonl(parent transcripts only; subagent transcripts undersubagents/are evidence, not primary scope). - Default scope: the most recent parent transcript with
lastWriteTimeolder than now (i.e. the just-finished session). - Optional flags (passed from
/pn-retro):--since=<YYYY-MM-DD>,--scope=session|recent|all.recent= last 5 parent transcripts;all= every parent transcript modified since the last retro report.
Transcripts are JSONL with {role: "user" | "assistant", message: {content: [...]}}. Assistant prose is stripped ("[REDACTED]"); only user text and tool calls (name + input) are reliably present. Classification must work from (user_text, tool_calls, tool_outputs) alone.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 182 lines · 85 tokens per session scan A bcd62a9c56a8
pn-session-retro is a skill published in the GitHub repository perniemann/pnCore (0 stars, last pushed 3d ago), licensed MIT. It adds 85 tokens to every session and 3,713 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
error-handling
Apply error handling and recovery patterns in JavaScript/TypeScript or Node.js. Use when implementing error handling, retry logic, or when the user mentions domain errors, error recovery, error escalation.
t-800-fix-pack
A defined format and workflow for preparing a fix pack, a document that lists approved changes to an existing artifact.
cocoreview
CocoReview — structured code review with six-severity findings vocabulary, progressive disclosure architecture, and universal anti-pattern baseline. Invoked via $review [file] [--complexity] [--security] [--architecture] [--language ].
firebase-crashlytics
Use when implementing crash reporting, capturing fatal/non-fatal errors, recording isolate/async exceptions, customizing reports, or uploading obfuscated symbols.
flutter-errors
Use when hitting layout errors (RenderFlex overflow, unbounded constraints, RenderBox not laid out), scroll errors, or setState-during-build errors.
identify-key-failures
Synthesize the common stuck points across failed decomposition plans. Use when the current batch of decomposition plans has failed — whether they failed already at direct proving or only after further attempts.