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 radiantlogicinc/fastworkflow --skill optimize-workflow-with-feedbackgit clone --depth 1 https://github.com/radiantlogicinc/fastworkflowWrote 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/radiantlogicinc/fastworkflow/optimize-workflow-with-feedback)<a href="https://agentmods.dev/skills/radiantlogicinc/fastworkflow/optimize-workflow-with-feedback"><img src="https://agentmods.dev/badge/skills/radiantlogicinc/fastworkflow/optimize-workflow-with-feedback/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/radiantlogicinc/fastworkflow/optimize-workflow-with-feedback"><img src="https://agentmods.dev/badge/skills/radiantlogicinc/fastworkflow/optimize-workflow-with-feedback.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 | $0.00066 | $0.01787 |
| Opus 5.5 | $0.00026 | $0.00715 |
| Sonnet 5.5 | $0.00013 | $0.00357 |
| Haiku 4.5 | $0.00007 | $0.00179 |
Grade A, and why
optimize-workflow-with-feedback 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 9d 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 — 143 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Optimize workflows using evidence and feedback
Start with a concrete user problem and the recorded behavior that demonstrates it. Deliver a trace-supported diagnosis, a scoped change, regression coverage and a comparison of observed outcomes. If new execution is not authorized or available, deliver the diagnosis, change and prepared comparison inputs, and clearly identify which claims remain unmeasured.
Find the evidence in run_chatbot
Select the application workflow and use benchmark → experiment → conversation on the left. For ordinary chats use ad-hoc conversations → UTC date → conversation. Inspect turns and then planning, execution steps and individual calls on the right. The right breadcrumb retains the entire drill-down path; the left tree ends at conversations.
Record the evidence store identity/path, benchmark version/digest, experiment ID, task ID, attempt, channel/conversation, turn key and relevant span IDs. Conversation IDs alone are not unique across channels or recordings. Follow the selected experiment's bound store; do not assume all results are in the currently selected workflow's default database.
Use debug-workflow-conversations for trace reads and stage-specific diagnosis. Before interpreting a missing call, inspect writer health and redaction: redacted, truncated, dropped and absent evidence are different conditions. Do not rerun a workflow merely to obtain an answer already present in its recording.
Use human feedback precisely
The Feedback box on a turn, phase, step or span appends a timestamped comment, categorized
with one of three categories and its two subcategories: Observations / Analysis
(observation, analysis), Conclusions (what_went_right, what_went_wrong) and
Recommendations (what_to_do, what_not_to_do). The comment itself is free text and no
heading in it is parsed. Comment on the narrowest component with recorded spans that explains the
problem; if the component has no span anchor, comment on the turn. Useful comments state:
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.
- 9d ago Changed · -3 lines 5efe4d874516
- 11d ago Changed · +6 lines d8c1ce241abb
- 1mo ago First seen · 140 lines · 66 tokens per session scan A 1f03aa4da1ca
optimize-workflow-with-feedback is a skill published in the GitHub repository radiantlogicinc/fastworkflow (52 stars, last pushed 5d ago), licensed Apache-2.0. It adds 66 tokens to every session and 1,787 once invoked, about $0.0003 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 2026-09-11.
Other skills, from other repositories
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.
eval-dataset-design
A guide to creating reliable test sets for AI agents. Each evaluation task includes the request, the simulated user's behaviour, a reset starting state, and a success check that can be independently verified.
tika-eval-compare
Compare extracts from two Tika builds over a corpus to detect regressions in content, encoding, exceptions, and embedded-document handling. Use for "compare before/after extracts", "eval this change against the corpus".
cua-driver
Use Cua Driver for desktop or browser tasks that are awkward or unavailable through Bash/APIs, or when the user explicitly wants GUI interaction: app testing, visual bug reproduction, form filling, calendar entry, screenshots, and demo recording. Also covers Cua setup; not OpenAI Codex Computer Use or web research.
jetson-validate-image
Use after jetson-flash-image to run static BSP checks, on-target smoke/regression tests on a flashed DUT, or both. Not for build or flash steps. Triggers: validate bsp, on-target validation.
atmos-validation
Validate Atmos projects, components, arbitrary JSON Schema inputs, EditorConfig, and GitHub Actions; use affected-file selection and native CI annotations.