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 event4u-app/agent-config --skill perf-feedback-craftgit clone --depth 1 https://github.com/event4u-app/agent-configWrote 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/event4u-app/agent-config/perf-feedback-craft)<a href="https://agentmods.dev/skills/event4u-app/agent-config/perf-feedback-craft"><img src="https://agentmods.dev/badge/skills/event4u-app/agent-config/perf-feedback-craft/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/event4u-app/agent-config/perf-feedback-craft"><img src="https://agentmods.dev/badge/skills/event4u-app/agent-config/perf-feedback-craft.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 | $0.00048 | $0.02726 |
| Opus 5 | $0.00024 | $0.01363 |
| Sonnet 5 | $0.00010 | $0.00545 |
| Haiku 4.5 | $0.00005 | $0.00273 |
Grade A, and why
perf-feedback-craft 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 8d 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 — 166 lines — stays where its author put it; the contents beside it link to each section on GitHub.
perf-feedback-craft
When to use
- A specific feedback conversation is upcoming (1:1, 30 / 60 / 90 check, mid-cycle review, end-of-cycle review, corrective conversation) and the question is what shape this conversation should take.
- A team-wide feedback cadence is being designed or audited and the question is which signals get surfaced, when, and through what channel.
- A growth conversation is being confused with a corrective conversation (or vice versa) and someone needs to separate them before the next exchange.
Do NOT use as a comp-decision surface (route to Q2 comp-banding; feedback informs comp, doesn't substitute for it), as a hiring-loop / calibration-design skill (route to S2 hiring-loop-design), or for performance-review-software configuration.
Cognition cluster
- Mental model 1 — First principles. Strip feedback to: what observation, about what behavior, with what impact, requesting what change? Most feedback fails because it skips the observation (jumps to interpretation) or skips the impact (assumes it's obvious). See
mental-models.md§ 1. - Mental model — Ladder of inference. Behavior observed → data selected → meaning inferred → assumptions made → conclusions drawn → beliefs adopted → action taken. Most feedback exchanges fail because giver and receiver are on different rungs. Naming the rung you're on (and inviting the other party to do the same) is the single highest-leverage feedback skill.
- Mental model 28 — Inversion. "What would make this feedback land as an attack instead of a gift?" — usually: public delivery, surprise (no prior signal), interpretation-as-fact, no specific request, no listening turn. Inversion surfaces the four canonical mis-deliveries.
- Mental model 21 — Second-order thinking. Feedback ripples. Praise in public, correction in private — the inverse damages both giver and receiver. A single mishandled feedback exchange damages trust for 6+ months; a single well-handled one banks trust that compounds.
- Context-spine slots. Read org-stage for what feedback infrastructure exists (10-person: ad-hoc; 50-person: cadence emerging; 150+: documented system). Read customer-segment and product for what behavior matters (B2B-enterprise sales = stakeholder management; consumer = velocity; deep-domain = quality bar).
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.
- 8d ago First seen · 166 lines · 48 tokens per session scan A 46a2a5061fb4
perf-feedback-craft is a skill published in the GitHub repository event4u-app/agent-config (10 stars, last pushed today), licensed MIT. It adds 48 tokens to every session and 2,726 once invoked, about $0.0002 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 2026-09-04.
Other skills, from other repositories
bazel-build-optimization
Optimize Bazel builds for large-scale monorepos. Use when configuring Bazel, implementing remote execution, or optimizing build performance for enterprise codebases.
sdd-tasks
Break an SDD change into implementation tasks. Trigger: orchestrator launches task planning for a change.
contact-cache
Track all identified/contacted people across strategies. CSV-backed contact database with dedup by LinkedIn URL or email. Prevents duplicate outreach when running strategies on a recurring cadence.
pad
Talk to your project. Natural-language project management — create items, check status, plan work, brainstorm ideas, and more.
pm-arch
Skill "pm-arch" from wei63w/pm-manager, covering user input, outline, done when and shared workflow (all /pm- commands).
pm-discover
Skill "pm-discover" from wei63w/pm-manager, covering user input, outline and shared workflow.