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/edobreque/clens/session-comparenpx skills add edobreque/clens --skill session-comparegit clone --depth 1 https://github.com/edobreque/clensWrote 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/edobreque/clens/session-compare)<a href="https://agentmods.dev/skills/edobreque/clens/session-compare"><img src="https://agentmods.dev/badge/skills/edobreque/clens/session-compare.svg" alt="Measured on agentmods" 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 | $0.00013 | $0.01328 |
| Opus 5 | $0.00006 | $0.00664 |
| Sonnet 5 | $0.00003 | $0.00266 |
| Haiku 4.5 | $0.00001 | $0.00133 |
Grade A, and why
session-compare 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 4d 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.
This is a copy
100% identical to session-compare — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
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.
Session Compare
Compare two distilled clens sessions side-by-side to identify performance differences and determine which approach was more effective.
Variables
TARGET: $ARGUMENTS
Instructions
1. Resolve the Sessions
- If TARGET contains two session IDs (space-separated), use those
- If TARGET is
--last 2or--last, find the two most recent distilled sessions:- List files in
.clens/distilled/sorted by modification time - Use the two most recently modified
.jsonfiles
- List files in
- If TARGET contains only one session ID, use it as Session A and the most recent other session as Session B
- If fewer than 2 distilled sessions exist, tell the user:
Need at least 2 distilled sessions to compare. Run
clens distillon more sessions first.
Label the older session as Session A and the newer as Session B.
2. Read Both Sessions
- Read both distilled JSON files from
.clens/distilled/{sessionId}.json - Parse both as
DistilledSessionobjects - Load the
session-analysisskill for schema and interpretation context
3. Generate the Comparison Report
Use the following template. Compute all deltas and provide analysis.
Session Comparison: {session_a_id first 8 chars} vs {session_b_id first 8 chars}
Metrics Comparison
| Metric | Session A | Session B | Delta | Better |
|---|---|---|---|---|
| Duration | {a.summary.key_metrics.duration_human} | {b.summary.key_metrics.duration_human} | {difference} | {arrow: shorter is better} |
| Tool calls | {a.stats.tool_call_count} | {b.stats.tool_call_count} | {+/- difference} | {fewer is generally better} |
| Failures | {a.stats.failure_count} | {b.stats.failure_count} | {+/- difference} | {fewer is better} |
| Failure rate | {a.stats.failure_rate as %}% | {b.stats.failure_rate as %}% | {+/- difference}pp | {lower is better} |
| Estimated cost | ${a cost} | ${b cost} | {+/- difference} | {lower is better} |
| Backtracks | {a.backtracks.length} | {b.backtracks.length} | {+/- difference} | {fewer is better} |
| Files touched | {a unique_files count} | {b unique_files count} | {+/- difference} | {context-dependent} |
| Files modified | {a.summary.key_metrics.files_modified} | {b.summary.key_metrics.files_modified} | {+/- difference} | {context-dependent} |
| Thinking blocks | {a.reasoning.length} | {b.reasoning.length} | {+/- difference} | {context-dependent} |
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.
- 4d ago First seen · 143 lines · 13 tokens per session scan A 446c1a56a6ab
session-compare is a skill published in the GitHub repository edobreque/clens (2 stars, last pushed yesterday), licensed MIT. It adds 13 tokens to every session and 1,328 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to session-compare, differing in 0 lines, and is treated as a copy.
Other skills, from other repositories
backtrack-analysis
Deep-dive into backtracking patterns from distilled clens sessions.
journey-report
Analyze cross-session development lifecycle and generate a journey report.
session-compare
Compare two distilled clens sessions side-by-side to identify performance differences.
session-report
Analyze a distilled clens session and generate a structured performance report.
session-analysis
Schema knowledge and interpretation guide for clens distilled session data. Auto-loads when reading .clens/distilled/ files.
api-endpoints
Adding or changing API operations in @repo/operations. One source of truth (defineOperation + a Zod schema) becomes an HTTP endpoint, an OpenAPI operation, an MCP tool, TS + Python SDK methods, a latitude CLI command, and an in-process agent tool — descriptions and contracts must be written with all of these readers…