So what to make of the denials? In order to put these in context, you need to understand an old Washington trick--the nondenial denial. The pattern is this: A news article alleges that some category of activity occurred. The "nondenial denial" does not categorically deny that something with in that category occurred, but instead denies only that specific types of activity within that category occurred. By inference, there is an admission (or at least not a denial) that something in that category occurred.
Here, there has not been a categorical denial that classified information was shared. Instead, we have more specific denials: “At no time were any intelligence sources or methods discussed, and no military operations were disclosed that were not already known publicly.” From this you can surmise that highly classified information was shared--just not intelligence sources or methods or secret military operations. From this, I surmise that the Washington Post article accurately reports that Trump "did not reveal the specific intelligence-gathering method, but he described how the Islamic State was pursuing elements of a specific plot and how much harm such an attack could cause under varying circumstances. Most alarmingly, officials said, Trump revealed the city in the Islamic State’s territory where the U.S. intelligence partner detected the threat." And Trump has since tweeted that he has the right to disclose classified information.
The calculation that needed to be made is whether the risks of disclosure are outweighed by the benefits of disclosure. Here the downside risks are enormous. the Washington Post article explains that the source of the information was a foreign intelligence agency that was itself highly classified (and not even disclosed to our allies). It was "code-word" information, which means that access was tightly controlled even among those with the highest security clearances. (My friend Jonathan Lee has a great primer on the highly classified nature of this information here.) The downside risk is that in disclosing both the existence of the plot and the city where the foreign partner detected this information, the Russians could deduce the source of the information and surmise that there is a human agent or technical means in that City.
The downsides here include the lack of future sharing by the foreign partner and the increased danger of exposure of the source. As the article explains, "The identification of the location was seen as particularly problematic, officials said, because Russia could use that detail to help identify the U.S. ally or intelligence capability involved. Officials said the capability could be useful for other purposes, possibly providing intelligence on Russia’s presence in Syria. Moscow would be keenly interested in identifying that source and perhaps disrupting it."
So when would the risk be worth taking? If the plot was directed toward Russia, the intelligence might be "actionable" in Russian hands, and therefore worth the downside risks. If, however, as the article suggests, Trump was merely bragging about what he knew, disclosure looks reckless.
To me the most damning evidence that this was not a calculated disclosure is the fact that the White House only discussed the disclosure with the intelligence community after the meeting. This suggests that there was not a pro/con vetting of the disclosure by those with knowledge of the situation.
So what should be done? I think the House and Senate Intelligence Communities are the logical places to get to ground truth. Sadly, given the highly classified nature of the intelligence here, their conclusions might not be made public. Still, they might be able to at least give a judgment about whether the benefits here were worth the large risks.