Chalk Labs
Need help deploying benchmark-integrity?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/codesbyfebin-benchmark-integrity
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: productivity
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/benchmark-integrity
What's inside
- Skill name
benchmark-integrity- Description
- Benchmark Integrity workflow for coding agents. Use when a new run is compared with the frozen board.
- License
- MIT
- Instructions
- 24 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Benchmark Integrity ## Workflow 1. Identify the requested outcome and the supplied evidence. 2. Require matching case, rubric, model, protocol, and coverage identities; never alter history. 3. Verify the result against the explicit user contract. 4. Report only completed work as completed; name one next action if work remains. ## Example Request: Did this candidate beat the snapshot? Response pattern: Check comparability and release criteria before making a ranking claim. ## Guardrails Preserve facts, uncertainty, safety requirements, and user-requested detail. Do not infer a diagnosis from preferences. Do not claim tools ran without execution evidence. This new skill is unbenchmarked; the frozen board remains held. ## Deterministic helper Use `compare_protocol` in `src/solvers.py` from the repository root. Read its catalog entry in `data/functions.json` for input and output examples. The helper transforms supplied data; it does not execute an agent task or replace a judge.
From CodesbyFebin/adhd-agent-harness, MIT.
More from CodesbyFebin/adhd-agent-harness
- 0.564score
agent-owned-work
CodesbyFebin
Agent Owned Work workflow for coding agents.
SkillAFreshness: AClaude Code +9 - 0.564score
bounded-next-action
CodesbyFebin
Bounded Next Action workflow for coding agents.
SkillAFreshness: AClaude Code +9 - 0.564score
decision-support
CodesbyFebin
Decision Support workflow for coding agents.
SkillAFreshness: AClaude Code +9 - 0.564score
evidence-aware-errors
CodesbyFebin
Evidence Aware Errors workflow for coding agents.
SkillAFreshness: AClaude Code +9 - 0.564score
non-diagnostic-boundary
CodesbyFebin
Non Diagnostic Boundary workflow for coding agents.
SkillAFreshness: AClaude Code +9 - 0.564score
output-contract
CodesbyFebin
Output Contract workflow for coding agents.
SkillAFreshness: AClaude Code +9 - 0.564score
partial-success
CodesbyFebin
Partial Success workflow for coding agents.
SkillAFreshness: AClaude Code +9 - 0.564score
progress-state
CodesbyFebin
Progress State workflow for coding agents.
SkillAFreshness: AClaude Code +9 - 0.564score
release-gate
CodesbyFebin
Release Gate workflow for coding agents.
SkillAFreshness: AClaude Code +9 - 0.564score
resolve-ambiguity
CodesbyFebin
Resolve Ambiguity workflow for coding agents.
SkillAFreshness: AClaude Code +9 - 0.564score
resume-context
CodesbyFebin
Resume Context workflow for coding agents.
SkillAFreshness: AClaude Code +9 - 0.564score
rollback-planning
CodesbyFebin
Rollback Planning workflow for coding agents.
SkillAFreshness: AClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 4, 2026 (26 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from CodesbyFebin
- 0.570score
linkedin-cta-lab
CodesbyFebin
Write three closing lines for an existing draft and score them.
SkillAFreshness: AClaude Code +9 - 0.570score
linkedin-connection-note
CodesbyFebin
Draft one connection note or follow-up for a named person.
SkillAFreshness: AClaude Code +9 - 0.570score
linkedin-analytics-review
CodesbyFebin
Read pasted LinkedIn analytics and say what to repeat.
SkillAFreshness: AClaude Code +9 - 0.570score
linkedin-carousel-writer
CodesbyFebin
Turn one claim into a LinkedIn document or carousel script.
SkillAFreshness: AClaude Code +9 - 0.567score
linkedin-case-receipt
CodesbyFebin
Draft a proof post from a build log: what shipped, what failed, what the log showed.
SkillAFreshness: AClaude Code +9 - 0.566score
linkedin-approval-gate
CodesbyFebin
Run the pre-publish gate on a draft before any post, comment, reply, invite, poll, or caption goes out.
SkillAFreshness: AClaude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get benchmark-integrity running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
benchmark-integrity skill. Agent Mart. https://agentmart.dev/skills/codesbyfebin-benchmark-integrity (last updated 2026-10-24)