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 skills/ilyagorsky/memory-toolkit/reflectnpx skills add IlyaGorsky/memory-toolkit --skill reflectgit clone --depth 1 https://github.com/IlyaGorsky/memory-toolkitWrote 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/ilyagorsky/memory-toolkit/reflect)<a href="https://agentmods.dev/skills/ilyagorsky/memory-toolkit/reflect"><img src="https://agentmods.dev/badge/skills/ilyagorsky/memory-toolkit/reflect.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 | $0.00016 | $0.00981 |
| Opus 5 | $0.00008 | $0.00491 |
| Sonnet 5 | $0.00003 | $0.00196 |
| Haiku 4.5 | $0.00002 | $0.00098 |
Grade A, and why
reflect scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 129 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reflect — session reflection
Analyze current session and propose additions to the project backlog.
1. Find backlog target
Check memory for saved preference:
node "$MEM" --dir="$MEM_DIR" search "backlog_target"
If no saved preference — detect automatically:
- Check common locations:
backlog.md,TODO.md,TODO,docs/backlog.md - Check if project uses GitHub Issues:
gh issue list --limit 1 2>/dev/null - If nothing found — ask:
Where do you track backlog?
1. backlog.md (will create)
2. TODO.md
3. GitHub Issues (gh issue create)
4. Other — specify path or tool
Save preference:
node "$MEM" --dir="$MEM_DIR" note "CONFIG: backlog_target=<chosen path or github-issues>"
Read current backlog to avoid duplicates.
1b. Collect watcher auto-findings
The session watcher captures decisions, plans, corrections, and phase changes in real time. Review them before analyzing manually — many candidates for backlog / DOC notes are already surfaced:
node "$MEM" --dir="$MEM_DIR" findings
Output groups entries by type (decision, plan, correction, phase). Treat them as candidates, not authoritative:
- correction — strong candidate for
feedback/memory - decision — candidate for
decisions/or backlog - plan — candidate for backlog entry
- phase — context only, usually skip
User confirms which findings promote to backlog / memory in step 3.
2. Analyze session
Review what happened and find:
- Workarounds — bugs/problems worked around, not fixed
- Repetitions — pattern that appeared >1 time, asks for automation
- Gaps — skill/rule that was missing, had to do manually
- Insights — decisions worth scaling to other areas
- Done — what from backlog was completed this session
DOC classification check
For each insight, ask: "Should any contributor to this project know this?" If yes — it's a DOC, not just a backlog item or feedback. Save it immediately:
node "$MEM" --dir="$MEM_DIR" note "DOC: <domain> — <insight>"
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.
- 4d ago First seen · 129 lines · 16 tokens per session scan A 9754ddc5db3e
reflect is a skill published in the GitHub repository IlyaGorsky/memory-toolkit (13 stars, last pushed 4mo ago), licensed MIT. It adds 16 tokens to every session and 981 once invoked, about $0.0001 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 2026-08-30.
Other skills, from other repositories
project-state-governor
Maintain evidence-backed project state across conversations, branches, reviews, and research cycles. Use when an AI coding agent must reconstruct or consolidate fragmented project documentation, reconcile conflicting history against code/tests/contracts/Git, classify active and completed work, verify completion…
status
Show what structured context exists and what's missing. Overview of bundles, SCDs, coverage across all 11 concerns, and compilation sync status.
openspec-bulk-apply-change
Use when multiple active OpenSpec changes should be applied concurrently in isolated worktrees with delegated verification and no merge.
plan
Epic decomposition into trackable, right-sized tasks. Three modes — audit-aware (codebase-audit reports), workflow-audit-aware (handoff.yaml with pre-rated findings), standalone (from scratch). Light convention scanning for projects without CLAUDE.md.
gh-issue-scheduler
Finds all open GitHub issues that haven't been replied to by the owner, summarizes them, and generates a solution plan. Use when the user wants to audit pending tasks or plan maintenance work.
triage
Apply a formal state machine to issues — assign category (bug/enhancement/question/spike) and state (needs-triage → needs-info → ready-for-agent → ready-for-human → wontfix). Issues marked ready-for-agent become inputs to supergraph:plan. Use when processing a backlog, reviewing new issues, or preparing work for…