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/reflexioai/claude-smart/update-prnpx skills add ReflexioAI/claude-smart --skill update-prgit clone --depth 1 https://github.com/ReflexioAI/claude-smartWhat 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 | $0.00071 | $0.01608 |
| Opus 5 | $0.00036 | $0.00804 |
| Sonnet 5 | $0.00014 | $0.00322 |
| Haiku 4.5 | $0.00007 | $0.00161 |
Grade A, and why
update-pr scanned grade A with 1 finding 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
<How the changes were verified — manual testing steps, automated tests run, curl commands, etc.> Copies of this mod
1 near-identical copy found in the catalogue:
- update-pr — 100% identical, 0 lines differ
How it starts
The opening of the file, as written. The whole thing — 128 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PR Updater
Push follow-up changes to an existing PR and update its description to reflect the new work.
Workflow
Step 1: Pre-flight Checks
- Verify GitHub CLI authentication — run
gh api user --jq '.login'to confirm the CLI is authenticated. If this fails, the user needs to rungh auth loginfirst. - Find the existing PR — run
gh pr view --json number,title,body,url,baseRefName,stateto find the PR for the current branch. If no PR exists, abort and suggest using/create-prinstead. Check thestatefield — if it is notOPEN, abort and inform the user that the PR is already merged or closed. - Verify clean git state — run
git statusto ensure no uncommitted changes. If there are uncommitted changes, run the/commitskill first to commit them. - Sync submodules — check if
.gitmodulesexists first. If it does, rungit submodule update --init --recursive. If not, skip this step.
Step 2: Sync with Base Branch
Ensure the feature branch is up-to-date with the base branch to avoid merge conflicts:
- Fetch latest remote — run
git fetch origin <base-branch>(use the base branch from the PR metadata) - Check for divergence — run
git log HEAD..origin/<base-branch> --onelineto see if the base branch has new commits - Rebase if needed — if there are new commits on the base branch:
a. Run
git rebase origin/<base-branch>b. Resolve conflicts — if the rebase hits conflicts:- Read the conflicting files to understand both sides
- Surface the conflict to the user — show both sides and ask which version to keep. Do not silently resolve conflicts.
- Stage resolved files:
git add <file> - Continue:
git rebase --continue - Repeat until rebase completes
- Verify clean state — run
git statusto confirm no unresolved conflicts remain - Push the branch — run
git push. If the rebase changed history, usegit push --force-with-leaseinstead. If--force-with-leasefails, it means someone else has pushed to this branch. Fetch the remote branch (git fetch origin <branch>), inspect the divergence (git log HEAD..origin/<branch> --oneline), and ask the user how to proceed — they may need to integrate the other contributor's changes first.
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 · 128 lines · 71 tokens per session scan A a5deb237bade
update-pr is a skill published in the GitHub repository ReflexioAI/claude-smart (774 stars, last pushed 3d ago), licensed Apache-2.0. It adds 71 tokens to every session and 1,608 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
reflexio-embedded
Captures user facts and procedural corrections into .reflexio/ so the agent learns across sessions. Use when: (1) user states a preference, fact, config, or constraint; (2) user corrects the agent and confirms the fix with an explicit 'good'/'perfect' or by moving on without re-correcting for 1-2 turns; (3) at start…
reflexio-consolidate
Run a full-sweep consolidation over all .reflexio/ files — TTL sweep + n-way cluster merge. Use when the user asks to 'clean up reflexio', 'consolidate memory', 'deduplicate playbooks', or suspects drift across sessions.
reflexio
When to use: Always active. This skill is the contract between you and the reflexio cross-session memory plugin.
learn
When to use: The user wants immediate extraction of skills/preferences from the current session — e.g., after a major correction, before context compaction, or to test the loop.
clear-all
When to use: The user explicitly asks to delete all locally-stored skills/preferences. This is destructive and unrecoverable.
show
When to use: The user wants to see what reflexio currently knows about this project.