- Home
- Skills
- Marketing & SEO
- Post Performance Debrief
Works with the AI Agents you already use
Post Performance Debrief
Turn a published post and its metrics into an evidence-led debrief and three controlled next tests.
Free
Post Performance Debrief
Example session with this skill installed
Platform: LinkedIn text post. Goal: newsletter clicks. Post: “Most creators do not need more ideas. They need a better way to learn from the posts they already published…” Seven-day metrics: 18,400 impressions; 312 reactions; 41 comments; 22 reposts; 96 saves; 138 profile visits; 27 link clicks. Comparable baseline: median of the previous five LinkedIn text posts at seven days: 12,200 impressions; 25 link clicks.
- Read your context and instructions
- Compiled the post performance debrief
Executive read
• Observed: impressions are above the supplied comparable baseline.
• Derived: link CTR by impressions = 27 / 18,400 × 100 = 0.15%.
• Hypothesis — Plausible: the post earned attention and saves, but the CTA bridge may not make the newsletter payoff specific enough.
Next test 1
Changed variable: CTA specificity only. Control: current closing sentence. Test: name the exact newsletter outcome while keeping topic, hook, length, and format stable. Primary metric: link clicks at seven days. Support: directional lift versus the supplied comparable baseline without a material collapse in saves.
The full debrief also includes data quality, content mechanics, the performance fingerprint, ranked hypotheses, three test cards, Preserve / Change / Stop guidance, and a causality note.
Connects securely to your tools. The creator never sees your data.
About this skill
Stop guessing why a post worked—or did not
Post Performance Debrief turns the content and metrics you already have into a structured learning loop. It separates what the data actually shows from what is merely plausible, calculates only defensible rates, and gives you three controlled next-post tests.
What you get
- A data-quality audit covering metric labels, reporting windows, denominators, and missing evidence.
- Clearly separated Observed, Derived, Hypothesis, and Unknown findings.
- Safe rate calculations with formulas and named denominators.
- A content-mechanics review of hook, structure, proof, pacing, CTA, and format fit.
- A performance fingerprint across attention, retention, resonance, distribution, action, and evidence coverage.
- Ranked explanations with evidence for and against each one.
- Exactly three controlled experiments that change one primary variable at a time.
- A concise Preserve / Change / Stop summary.
Built for creators who want useful learning, not algorithm folklore
The skill never invents metrics, universal benchmarks, audience reactions, or platform causes. Missing evidence stays missing. A single post produces provisional hypotheses—not fake certainty or growth guarantees.
Best inputs
The published caption, script, transcript, or post text; platform and format; at least one observed metric; and optional goal, reporting window, comparable baseline, and context.
Known limitations
This skill does not connect to social accounts, scrape private dashboards, fetch live analytics, create posts from scratch, analyze paid campaigns, guarantee reach, or prove algorithmic causality. Cross-platform metrics are analyzed separately unless their definitions and conditions are genuinely comparable.
Requires a supplied post, platform or format, and at least one observed metric. Cannot scrape accounts, fetch live analytics, analyze paid campaigns, provide account-wide strategy, guarantee reach, or prove algorithmic causality. Cross-platform metrics are analyzed separately unless their definitions and conditions are comparable.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
This skill has been used 9 times.
Trust & safety
Security scan
Security scan passed 3 months ago
- Free with an Agensi account
Permissions
No special permissions declared or detected