The readme file itself should contain:

1. Short overview of programme - a few lines describing 
 (a) Main aim and technical set-up, e.g. 
      1" long slit spectroscopy in J-band of very bright stars, 
      multi-colour imaging of AGN host galaxies, 
      deep MOS spectroscopy of very faint galaxies in 2 bands, 
      etc.
 (b) What are the most important aspects for success of the project, 
     e.g. signal-to-noise, high spectral resolution, spatial resolution, 
     photometry, saving frames individually rather than summing, 
     high quality calibration, etc.

2. Preferred conditions: airmass, seeing, lunar phase, thin clouds or clear, etc

3. (a) Minimum acceptable atmospheric conditions 
   (b) Alternative observing set-ups (provide alternative scripts 
       as well, observers at the telescope will not modify 
       observing scripts)
       e.g. 2" rather than 1" slit for poor seeing, increased
       exposure times for poor transparency, etc.

4. List of submitted observing scripts with
   (a) target ID (matching that in the script name & finding chart)
   (b) target coordinates
   (c) target priority
   (d) total integration times including overheads
5. List of calibrators
   (e) list of photometric/telluric calibration stars or other calibrators
6. List of submitted masks

5. time cadence
   (a) execution times for time critical programs
   (b) any request to execute observing scripts in a particular order, 
   e.g. telluric before or after science targets, ordered list of scripts, etc. 

6. Any other critical issues such as special calibration requests, 
   or things the PI thinks could help the observers to obtain the best possible data.

7. Detailed contact information before and during a run for eavesdropping