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 agentmods add instructions/rjmurillo/ai-agents/retrosgit clone --depth 1 https://github.com/rjmurillo/ai-agentsWrote 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/instructions/rjmurillo/ai-agents/retros)<a href="https://agentmods.dev/instructions/rjmurillo/ai-agents/retros"><img src="https://agentmods.dev/badge/instructions/rjmurillo/ai-agents/retros.svg" alt="Measured on agentmods" 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 | $0.00407 | $0.00407 |
| Opus 5 | $0.00204 | $0.00204 |
| Sonnet 5 | $0.00081 | $0.00081 |
| Haiku 4.5 | $0.00041 | $0.00041 |
Grade A, and why
ai-agents retros.instructions.md 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 yesterday.
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.
What it actually says
Retrospective File Rules
Retrospectives exist so the same failure does not happen twice. They feed future governance and instruction changes, so the format matters.
MUST
- Filename convention. New retros MUST use
YYYY-MM-DD-<slug>.md(e.g.,2026-04-21-instruction-files-rollout.md). - Failure mode classification. Each retro MUST classify the failure against
.agents/governance/FAILURE-MODES.md. If no existing class matches, MUST propose a new class in a linked ADR. - Evidence. MUST include links to the offending commits, PRs, issues, or CI runs. No hand-waving.
- Remediation. MUST list concrete follow-up actions (governance change, ADR, instruction update, skill change) with owners or issues.
- No blame. MUST critique processes and artifacts, never individuals.
SHOULD
- Impact table. SHOULD include an
Impacttable with severity (High / Medium / Low) per affected area. - Root cause. SHOULD apply the five-whys or timeline analysis and write the root cause explicitly.
- Learning capture. SHOULD use the
reflectskill to extract patterns and persist HIGH-confidence learnings to Serena memory.
MUST NOT
- MUST NOT delete or edit landed retros to soften criticism. Corrections append a new section with a date and rationale.
- MUST NOT ship remediation commits without linking the retro in the commit body or PR.
References
.agents/governance/FAILURE-MODES.md. Failure mode taxonomy.claude/skills/reflect/SKILL.md. Learning-capture workflow.claude/skills/retro/. Weekly retrospective skill (if present).agents/retrospective/. Historical examples
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.
- yesterday First seen · 34 lines · 407 tokens per session scan A 60e991b1fc5f
ai-agents retros.instructions.md is an instructions file published in the GitHub repository rjmurillo/ai-agents (45 stars, last pushed yesterday), licensed MIT. It adds 407 tokens to every session, about $0.0020 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 instructions, from other repositories
abtars AGENTS.md
Instructions for aksika/abtars, covering agents.md, quick commands, architecture, model-call chokepoint: spin(spec) (#1271) and external dependency: abmind.
Alex_Skill_Mall falsifiability-deadlines.instructions.md
Every new or edited brain artefact (instruction / skill / prompt / agent) must declare a specific falsification deadline — a literal date, an observable event, or count+time bound — not 'after N passes' or 'when conditions warrant'.
norma CLAUDE.md
Instructions for yanlingLabs/norma, covering claude.md, what this is, commands, tests and swift.
Alex_Skill_Mall mall-maintenance-rules.instructions.md
Always-on routing for Mall maintenance work — fires the right Mall skill at the right moment. Distinguishes Mall-owned automation from out-of-scope editorial work.
Alex_Skill_Mall lint-discipline.instructions.md
Fix lint always — if I edited a file, I own its lint state on exit, even for pre-existing findings.
bed-of-roses CLAUDE.md
Instructions for SixHq/bed-of-roses, covering claude.md, project, commands, architecture and server/ — node runtime (the brain).