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 TeamRetroHQ/teamretro-skills --skill ai-session-retrogit clone --depth 1 https://github.com/TeamRetroHQ/teamretro-skillsWrote 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/teamretrohq/teamretro-skills/ai-session-retro)<a href="https://agentmods.dev/skills/teamretrohq/teamretro-skills/ai-session-retro"><img src="https://agentmods.dev/badge/skills/teamretrohq/teamretro-skills/ai-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/teamretrohq/teamretro-skills/ai-session-retro"><img src="https://agentmods.dev/badge/skills/teamretrohq/teamretro-skills/ai-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 | $0.00104 | $0.02746 |
| Opus 5 | $0.00052 | $0.01373 |
| Sonnet 5 | $0.00021 | $0.00549 |
| Haiku 4.5 | $0.00010 | $0.00275 |
Grade A, and why
ai-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 12d 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 — 111 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI Session Retro
At the end of a working session, produce an honest, evidence-based retrospective of that session and save it to a persistent log. Over time these entries accumulate into a dataset the team can analyze (see the companion ai-retro-brief skill).
One session's friction is noise; the same friction across eight sessions is a process problem. The log is only useful if entries are consistent (identical schema, so they're comparable) and honest (specific, evidence-cited observations, not vibes). Treat vagueness as a failure mode.
The root-cause vocabulary below is a synchronized copy; its canonical definition and change process live in the team's engineering docs (docs-first). Never alter labels ad hoc — vocabulary changes happen at a periodic review and arrive here as a version bump.
Skill invocation is probabilistic — capture fails silently when nothing triggers it. See SETUP.md for hook, command, and convention snippets that make the trigger deterministic.
The log
One file per entry: docs/ai-retros/entries/YYYY-MM-DD-<short-slug>.md, relative to the repo root. Create the directories if needed. Never edit or delete existing entries. If the slug would collide with an existing file, append -2.
The log lives at the work's review scope — the boundary where fixes land. For most dev work that's the repo: its docs, config, and conventions are repo-homed, so its friction log is too. For non-code work (an ads account, a support inbox, a client engagement), keep one log per scope in that workspace's document home. If there's no natural home, ask the user where the team's shared log should live — an entry written to a scratch directory dies with the session, which defeats the purpose.
Commit the new entry as its own commit (docs: add AI session retro entry). If you're on a feature branch or worktree, that's fine — the entry merges with the work it describes.
Step 1 — Review the actual session
Walk back through the conversation and collect concrete moments: instructions that were ambiguous and forced a guess, context or access that was missing, code, structures, or docs that made the work slower than it should have been, work redone after a correction, tool failures, and things that made the work fast. Every observation must trace to one of these moments — if you can't cite a moment, drop the observation.
What ships with it
6 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- evals/evals.json 2.9 KB
- evals/fixtures/existing-entries/docs/ai-retros/entries/2026-07-06-support-ticket-theme-analysis.md 790 B
- evals/fixtures/existing-entries/docs/ai-retros/entries/2026-07-09-onboarding-email-sequence.md 594 B
- evals/fixtures/existing-entries/transcript-2.md 1.1 KB
- evals/fixtures/transcript-1.md 1.6 KB
- SETUP.md 3.4 KB
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.
- 12d ago First seen · 111 lines · 104 tokens per session scan A 5f9023dc453c
ai-session-retro is a skill published in the GitHub repository TeamRetroHQ/teamretro-skills (9 stars, last pushed 1mo ago), licensed MIT. It adds 104 tokens to every session and 2,746 once invoked, about $0.0005 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 2026-08-31.
Other skills, from other repositories
token-budget-advisor
Enforces Law 2 (Plan Is Sacred) of the 7 Laws of AI Agent Discipline by making token-budget tradeoffs explicit before the response is composed. Offers the user an informed choice about how much response depth to consume before answering. Use this skill when the user explicitly wants to control response length, depth…
run-team-retrospective
Use when a team needs to reflect on a completed period, project, or significant event — to extract what worked, what didn't, and what specific changes to make before the next cycle, rather than repeating the same patterns.
founder-coach
Personal leadership development for founders and first-time CEOs. Use when a founder feels like the bottleneck, struggles to delegate, is burning out, or transitioning from IC to executive.
iterate-step-by-step
How to use the 4-Turn Conversational Scaffolding framework to build complex code, strategy, and content incrementally without token exhaustion or quality degradation.
knowledge-ops
Audit and repair an internal knowledge base — staleness, ownership gaps, orphans, duplication, and findability. Use when the wiki is untrusted, docs are out of date, nobody owns pages, or search returns the wrong answer.
apply-5s-methodology
Use when a physical workspace, shared digital environment, or shop floor has accumulated clutter, unclear organization, or inconsistent upkeep that slows work down or hides problems — apply the five sequential steps (sort, set in order, shine, standardize, sustain) in order, because skipping ahead to standardization…