Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/rjmurillo/ai-agentsnpx agentmods add skills/rjmurillo/ai-agents/memory-reflexionWrote 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/rjmurillo/ai-agents/memory-reflexion)<a href="https://agentmods.dev/skills/rjmurillo/ai-agents/memory-reflexion"><img src="https://agentmods.dev/badge/skills/rjmurillo/ai-agents/memory-reflexion/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/rjmurillo/ai-agents/memory-reflexion"><img src="https://agentmods.dev/badge/skills/rjmurillo/ai-agents/memory-reflexion.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00098 | $0.01872 |
| Opus 5.5 | $0.00039 | $0.00749 |
| Sonnet 5.5 | $0.00020 | $0.00374 |
| Haiku 4.5 | $0.00010 | $0.00187 |
Grade A, and why
memory-reflexion 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 1mo 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 — 198 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Memory Reflexion
Tier 2 episodic write operations: the reflexion write path extracted from the
memory router per ADR-063 (memory-skill decomposition). memory still routes
here; a caller that needs only to record a session outcome loads this sub-skill
instead of the full memory surface.
Reflexion is one operation: extract an episode from a completed session log.
The episode record is the durable artifact. It is committed to the repository
and can be queried through memory-search. No non-memory downstream consumer
reads it today.
[!NOTE] This sub-skill previously carried a second step that folded episodes into a derived causal graph (Tier 3). ADR-089 removed that layer: nothing read it, and its aggregated output was noise. Episodes are unaffected and remain the system of record. See ADR-089 for the evidence.
Triggers
Use this skill when the user says:
extract episode from sessionto capture a completed session as an episoderecord what happened this sessionfor the reflexion write pathre-extract this episodeto refresh a record over an existing episode file
Quick Start
# Extract an episode from a completed session log (Tier 2)
python3 .claude/skills/memory/scripts/extract_session_episode.py "<session-log-path>"
# Re-extract over an existing episode, merging rather than clobbering
python3 .claude/skills/memory/scripts/extract_session_episode.py "<session-log-path>" --preserve
The script is canonical and shared with the memory router; this sub-skill does
not reimplement it. It lives at
.claude/skills/memory/scripts/extract_session_episode.py.
Quick Reference
| Operation | Script | Key Parameters |
|---|---|---|
| Extract episode (Tier 2) | extract_session_episode.py |
session_log_path, --output-path, --force, --preserve |
--force and --preserve are mutually exclusive. --force overwrites an
existing episode file. --preserve merges a fresh extraction over the existing
record without dropping richer data already there.
What ships with it
1 file 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.
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.
- 1mo ago Changed · +5 lines ba85d90d733a
- 1mo ago Changed · +2 lines 484eb07d1b15
- 1mo ago Changed · +1 lines 5d6dd094befa
- 1mo ago Changed 6e9a2f329457
- 1mo ago Changed 9e89ccb6c526
- 1mo ago First seen · 190 lines · 98 tokens per session scan A 8d4af81f727f
memory-reflexion is a skill published in the GitHub repository rjmurillo/ai-agents (49 stars, last pushed today), licensed MIT. It adds 98 tokens to every session and 1,872 once invoked, about $0.0004 per session on Opus 5.5. 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
apex-context-management
UTILITY SKILL — Two-mode context-window management. RUNTIME: artifact compression (full/summarized/minimal) for orchestrator and codegen agents. AUDIT: post-mortem analysis of Copilot debug logs (token profiling, redundancy, hand-off gaps) for 11-Context Optimizer. WHEN: "context optimization", "token budget"…
install-gemini-code-review-action
Installs, configures, or upgrades Dazbo's PR code review & issue triage GitHub Action (gemini-review-action) in a target repository. Use when the user wants to set up, configure, or update automated AI code reviews, issue triaging, or Gemini model workflows for their GitHub repository.
memory
Read, write, and summarize persistent Office memory when durable behavioral guidance or project knowledge should survive across agent sessions.
apply-review-feedback
Classify and apply accepted PR/MR review feedback.
review-remote
Review open PR/MR and publish findings.
phoenix-review
Review completed work against the Intent Contract's acceptance criteria using objective evidence, not opinion — re-run every check, confirm no regressions, inspect the tamper-evident trace, and surface only real issues (bugs, unmet criteria, regressions). Use before shipping, after a feature is built, or when the user…